Google AutoML is a tool from Google that automates the process of developing machine learning algorithms for various tasks. It’s user-friendly, fairly simple to use, and cloud-based.

AutoML is implemented in five different applications:

• Vision
• Video Intelligence
• Natural Language
• Translation
• Tables

Neural Architecture Search implemented as a RNN that generates architectures encoded as a variable-length sequence of tokens. The sequence of tokens generated can be seen as a sequence of actions that should be done to generate the architecture.

Once an architecture is generated, the proposed model is constructed and trained to finally record the acquired accuracy. The RNN is trained used reinforcement learning with a policy that updates the parameters of the RNN so that it generates better architectures over time.

Google Cloud AutoML enables developers with limited machine learning expertise to train high-quality models specific to their business needs.
Click To Tweet

Cloud AutoML enables developers with limited machine learning expertise to train high-quality models specific to their business needs. It relies on Google’s state-of-the-art transfer learning and neural architecture search technology.

AutoML Workflow

The businesses can bring their dataset, identify labels, and set the button. The steps that take place between data acquisition and prediction is abstracted by the AutoML platform. The trained and optimized machine learning model predicts. In brief, most of the steps are handled behind the scene providing time and opportunity for businesses to stay focused on problem and solutions rather than process and workflow.

Many of the AutoML platforms support to export the trained model into mobile devices running in iOS and Android, enabling developers to integrate easily within their mobile devices. As they get exported into the Docker container, DevOps team can deploy them at scale and infer in production environments.

AutoML Vision

AutoML Vision enables you to train machine learning models to classify your images according to own defined labels.

• Train models from labeled images and evaluate their performance.
• Leverage a human labeling service for datasets with unlabeled images.
• Register trained models for serving through the AutoML API.

Cloud Vision includes several options to use in order to integrate machine learning vision models into  applications and web sites.

AutoML Vision Edge now allows also to export custom trained models.

• AutoML Vision Edge allows to train and deploy low-latency, high accuracy models optimized for edge devices.
• With Tensorflow Lite, Core ML, and container export formats, AutoML Vision Edge supports a variety of devices.
• Hardware architectures supported: Edge TPUs, ARM and NVIDIA.

Cloud AutoML Vision Object Detection enables developers to train custom machine learning models that are capable of detecting individual objects in a given image along with its bounding box and label.

Cloud Vision Product Search allows retailers to create products, each containing reference images that visually describe the product from a set of viewpoints. Retailers can then add these products to product sets. Currently Vision API Product Search supports the following product categories: homegoods, apparel, toys, packaged goods, and general .

To build an application on iOS or Android devices use AutoML Vision Edge in ML Kit. This solution is available via Firebase and offers an end-to-end development flow for creating and deploying custom models to mobile devices using ML Kit client libraries.

AutoML Video Intelligence

Video Intelligence includes several options for integrating machine learning video intelligence models into your applications and web sites.

AutoML Video Intelligence Classification enables you to train machine learning models to classify shots and segments in your videos according to your own defined labels.

AutoML Video Intelligence Object Tracking enables you to train machine learning models to detect and track multiple objects in shots and segments. You can use these models to track objects in your videos according to your own pre-defined, custom labels.

AutoML Natural Language

AutoML Natural Language enables building and deploying custom machine learning models that analyze documents, categorizing them, identify entities within them, or assessing attitudes within them.

If you don’t need a custom model solution, the Cloud Natural Language API provides content classification, entity and sentiment analysis, and more.

AutoML Natural Language enables you to build and deploy custom machine learning models that analyze documents, categorizing them, identify entities within them, or assessing attitudes within them.
Click To Tweet

AutoML Natural Language uses machine learning to analyze the structure and meaning of documents. You train a custom machine learning model to classify documents, extract information, or understand the sentiment of authors.

•  A classification model analyzes a document and returns a list of content categories that apply to the text found in the document.
•  An entity extraction model inspects a document for known entities referenced in the document and labels those entities in the text.
•  A sentiment analysis model inspects a document and identifies the prevailing emotional opinion within it, especially to determine a writer’s attitude as positive, negative, or neutral.

Basic classification, entity extraction, and sentiment analysis are available through the Cloud Natural Language API. AutoML Natural Language enables defining custom classification categories, entities, and sentiment scores that are relevant to your application.

AutoML Translation

AutoML Translation allows to create own, custom translation models so that translation queries return results specific to the domain.

AutoML Translation trains custom models using matching pairs of sentences in the source and target languages. It treats each sentence pair as an independent training item, without assuming any correlation between separate pairs.

The sentence pairs used to train your custom model must be in Tab-separated values (.tsv) or Translation Memory eXchange (.tmx) format. It is possible to batch multiple .tsv and .tmx files into a comma-separated values (.csv) file.

Translation Memory eXchange (TMX) is a standard XML format for providing source and target translation sentences. AutoML Translation supports input files in a format based on TMX version 1.4. This example illustrates the required structure:

AutoML Tables

AutoML Tables enables entire team of data scientists, analysts, and developers to automatically build and deploy state-of-the-art machine learning models on structured data at massively increased speed and scale. Transform your business by leveraging your enterprise data to tackle mission-critical tasks like supply chain management, fraud detection, lead conversion optimization, and increasing customer lifetime value.

AutoML Tables is a supervised learning service. This means that you train a machine learning model with example data. AutoML Tables uses tabular (structured) data to train a machine learning model to make predictions on new data. One column from the dataset, called the target, is what your model will learn to predict. Some number of the other data columns are inputs (called features) that the model will learn patterns from. You can use the same input features to build multiple kinds of models just by changing the target. From the email marketing example, this means you could build two models with the same input features: One model could predict a customer’s persona (a categorical target), and one could predict their monthly spending (a numerical target).

AutoML Tables enables entire team of data scientists, analysts, and developers to automatically build and deploy state-of-the-art machine learning models on structured data at massively increased speed and scale.
Click To Tweet

Google Auto-ML advantages

Google’s AutoML efforts aim to make ML more scalable and accelerate both research and industry applications. Initial efforts of neural architecture search have enabled breakthroughs in computer vision with NasNet, and evolutionary methods such as AmoebaNet and hardware-aware mobile vision architecture MNasNet further show the benefit of these learning-to-learn methods.

Recently a learning-based approach was applied to tabular data, creating a scalable end-to-end AutoML solution that meets three key criteria:

• Full automation: Data and computation resources are the only inputs, while a servable TensorFlow model is the output. The whole process requires no human intervention.
• Extensive coverage: The solution is applicable to the majority of arbitrary tasks in the tabular data domain.
• High quality: Models generated by AutoML has comparable quality to models manually crafted by top ML experts.

About Google

Google LLC is an American multinational technology company that specializes in Internet-related services and products, which include online advertising technologies, a search engine, cloud computing, software, and hardware.The company’s rapid growth since incorporation has triggered a chain of products, acquisitions, and partnerships beyond Google’s core search engine (Google Search). It offers services designed for work and productivity (Google Docs, Google Sheets, and Google Slides), email (Gmail), scheduling and time management (Google Calendar), cloud storage (Google Drive), instant messaging and video chat (Duo, Hangouts, Meet), language translation (Google Translate), mapping and navigation (Google Maps, Waze, Google Earth, Street View), video sharing (YouTube), note-taking (Google Keep), and photo organizing and editing (Google Photos).

Read more about H2O Driverless AI, Data Science Machine, Azure AutoML and other automated machine learning software.

Subscribe to get latest updates:

https://twitter.com/sakalosch

The post Google AutoML appeared first on Quantum: Machine Learning & Analytics.

Azure Auto-ML

Azure Automated Machine Learning  –  cloud-based environment for  training, deploying, automating, managing and tracking ML models. Automated machine learning picks an algorithm and hyperparameters for you and generates a model ready for deployment.

Automated ML democratizes the machine learning model development process, and empowers its users, no matter their data science expertise, to identify an end-to-end machine learning pipeline for any problem.

Data scientists, analysts, and developers across industries can use automated ML to:

• Implement ML solutions without extensive programming knowledge
• Save time and resources
• Leverage data science best practices
• Provide agile problem-solving

Benefits for companies:

• Automatically build and deploy predictive models using the no-code UI or through a code-first notebooks experience.
• Increase productivity with easy data exploration and profiling and with intelligent feature engineering.
• Easily create accurate models customized to your data and refined by a wide array of algorithms and hyperparameters.
• Build responsible AI solutions with model interpretability, and fine-tune your models to improve accuracy.

Azure  AutoML: classify, regression, & forecast

Azure Automated Machine Learning could be applied to classifications, regressions and  forecasting. Classification is a common machine learning task. Azure Machine Learning offers featurizations specifically for these tasks, such as deep neural network text featurizers for classification. Common classification examples include fraud detection, handwriting recognition, and object detection.

Regression tasks are also a common supervised learning task. Azure Machine Learning offers featurizations specifically for these tasks. Different from classification where predicted output values are categorical, regression models predict numerical output values based on independent predictors. In regression, the objective is to help establish the relationship among those independent predictor variables by estimating how one variable impacts the others. For example, automobile price based on features like, gas mileage, safety rating, etc.

Time-series forecasting is an integral part of any business, whether it’s revenue, inventory, sales, or customer demand. Automated ML can be used to combine techniques and approaches and get a recommended, high-quality time-series forecast.

Automated ML can be used to time-series forecasting, which is an integral part of any business, whether it's revenue, inventory, sales, or customer demand.
Click To Tweet

An automated time-series experiment is treated as a multivariate regression problem. Past time-series values are “pivoted” to become additional dimensions for the regressor together with other predictors. This approach, unlike classical time series methods, has an advantage of naturally incorporating multiple contextual variables and their relationship to one another during training. Automated ML learns a single, but often internally branched model for all items in the dataset and prediction horizons. More data is thus available to estimate model parameters and generalization to unseen series becomes possible.

Advanced forecasting configuration includes:

• holiday detection and featurization
• time-series and DNN learners (Auto-ARIMA, Prophet, ForecastTCN)
• many models support through grouping
• rolling-origin cross validation
• configurable lags

AutoML workflow

During training, Azure Machine Learning creates a number of pipelines in parallel that try different algorithms and parameters for you. The service iterates through ML algorithms paired with feature selections, where each iteration produces a model with a training score. The higher the score, the better the model is considered to “fit” the data. It will stop once it hits the exit criteria defined in the experiment.

Using Azure Machine Learning, design and run automated ML training experiments with these steps:

1. Identify the ML problem to be solved: classification, forecasting, or regression
2. Choose whether to use the Python SDK or the studio web experience:
3. Specify the source and format of the labeled training data: Numpy arrays or Pandas dataframe
4. Configure the compute target for model training, such as your local computer, Azure Machine Learning Computes, remote VMs, or Azure Databricks.
5. Configure the automated machine learning parameters that determine how many iterations over different models, hyperparameter settings, advanced preprocessing/featurization, and what metrics to look at when determining the best model.
6. Submit the training run.
7. Review the results

Model Configuration

There are several options that you can use to configure automated machine learning experiments.

• Configuration options available in automated machine learning:
• Select your experiment type: Classification, Regression, or Time Series Forecasting
• Data source, formats, and fetch data
• Choose your compute target: local or remote
• Automated machine learning experiment settings
• Run an automated machine learning experiment
• Explore model metrics
• Register and deploy model

Automated machine learning supports data that resides on your local desktop or in the cloud such as Azure Blob Storage. The data can be read into a Pandas DataFrame or an Azure Machine Learning TabularDataset.

Feature engineering

Feature engineering is the process of using domain knowledge of the data to create features that help ML algorithms learn better. In Azure Machine Learning, scaling and normalization techniques are applied to facilitate feature engineering. Collectively, these techniques and feature engineering are referred to as featurization.

Automated machine learning featurization steps (feature normalization, handling missing data, converting text to numeric, etc.) become part of the underlying model. When using the model for predictions, the same featurization steps applied during training are applied to your input data automatically.

In every automated machine learning experiment, data is automatically scaled or normalized to help algorithms perform well. During model training, one of the following scaling or normalization techniques will be applied to each model. Additional feature engineering techniques such as, encoding and transforms are also available.

Ensemble models

Automated machine learning supports ensemble models, which are enabled by default. Ensemble learning improves machine learning results and predictive performance by combining multiple models as opposed to using single models. The ensemble iterations appear as the final iterations of the run. Automated machine learning uses both voting and stacking ensemble methods for combining models:

• Voting: predicts based on the weighted average of predicted class probabilities (for classification tasks) or predicted regression targets (for regression tasks).
• Stacking: stacking combines heterogenous models and trains a meta-model based on the output from the individual models. The current default meta-models are LogisticRegression for classification tasks and ElasticNet for regression/forecasting tasks.

Azure Auto-ML supports ensemble models, which improve machine learning results and predictive performance by combining multiple models as opposed to using single models.
Click To Tweet

Many models

The Many Models Solution Accelerator (preview) builds on Azure Machine Learning and enables you to use automated ML to train, operate, and manage hundreds or even thousands of machine learning models.

For example, building a model for each instance or individual in the following scenarios can lead to improved results:

• Predicting sales for each individual store
• Predictive maintenance for hundreds of oil wells
• Tailoring an experience for individual users.

AutoML on Databricks

AutoML on Databricks automates Machine Learning pipelines from feature engineering, model search, hyperparameter tuning, and inference while providing data scientists with the flexibility and control they need.

Databricks automates various steps of the data science workflow including augmented data preparation, visualization, feature engineering, hyperparameter tuning, model search, and finally automatic model tracking, reproducibility, and deployment, through a combination of native product offerings, partnerships, and custom solutions for a fully controlled and transparent AutoML experience.

Benefits

• Scalability: Automatically scale up and down your workloads and speed up training time with out-of-the-box optimizations for the most popular ML frameworks.
• Control: Choose algorithms better suited for the task in either single node or multi-node environment, and limit the number of runs to keep costs down.
• Ease of use: Automatically log results with MLflow tracking and parallelize hyperparameter search with Hyperopt on Databricks.
• Unification: Run all AutoML steps on the same platform, from ETL to model training and inference, securely, collaboratively, and at scale.

Databricks automates various steps of the data science workflow including augmented data preparation, visualization, feature engineering, hyperparameter tuning, model search, and finally automatic model tracking, reproducibility, and…
Click To Tweet

Features

• MLflow Experiments Tracking
• Track, compare, and visualize hundreds of thousands of experiments using open source or Managed MLflow.
• Automated Hyperparameter Tuning for Distributed Machine Learning
• Deep integration with PySpark MLlib’s Cross Validation to automatically track MLlib experiments in MLflow.
• Automated Hyperparameter Tuning for Single-node Machine Learning
• Optimized and distributed hyperparameter search with enhanced Hyperopt and automated tracking to MLflow.
• Automated Model Search for Single-node Machine Learning
• Optimized and distributed conditional hyperparameter search with enhanced Hyperopt and automated tracking to MLflow.
• Databricks Labs AutoML Toolkit
• Automated end-to-end model building pipeline is available via Databricks Labs custom solutions.

AutoML & ONNX

With Azure Machine Learning,  use automated ML to build a Python model and have it converted to the ONNX format. Once the models are in the ONNX format, they can be run on a variety of platforms and devices.

The ONNX runtime also supports C#, so use the model built automatically in C# apps without any need for recoding or any of the network latencies that REST endpoints introduce.

About Microsoft Azure

Microsoft Azure  is a cloud computing service created by Microsoft for building, testing, deploying, and managing applications and services through Microsoft-managed data centers. It provides software as a service (SaaS), platform as a service (PaaS) and infrastructure as a service (IaaS) and supports many different programming languages, tools, and frameworks, including both Microsoft-specific and third-party software and systems.

Read more about automated machine learning here.

Here you can find inforamtion about Data Science Machine for automated machine learning from MIT.

Subscribe to get latest updates:

Tweets by sakalosch

The post Azure Auto-ML appeared first on Quantum: Machine Learning & Analytics.

H2O Driverless AI Platform

H2O Driverless AI is a high-performance, GPU-enabled computing platform for automatic development and rapid deployment of predictive analytics models. It reads tabular data from plain text sources and from a variety of external data sources, and it automates data visualization and the construction of predictive models.

H2O Driverless AI is a platform for data science and machine learning automation of the workflows such as feature engineering, model validation, model tuning, model selection and model deployment.
Click To Tweet

Driverless AI also includes Machine Learning Interpretability (MLI), which incorporates a number of contemporary approaches to increase the transparency and accountability of complex models by providing model results in a human-readable format.

Driverless AI targets business applications such as loss-given-default, probability of default, customer churn, campaign response, fraud detection, anti-money-laundering, demand forecasting, and predictive asset maintenance models.

Key features

Flexibility of Data and Deployment: Driverless AI works across a variety of data sources including Hadoop HDFS, Amazon S3, and more. Driverless AI can be deployed everywhere including all clouds (Microsoft Azure, AWS, Google Cloud) and on premises on any system, but it is ideally suited for systems with GPUs, including IBM Power 9 with GPUs built in.

NVIDIA GPU Acceleration: The platform is optimized to take advantage of GPU acceleration to achieve up to 40X speedups for automatic machine learning. It includes multi-GPU algorithms for XGBoost, GLM, K-Means, and more. GPUs allow for thousands of iterations of model features and optimizations.

Automatic Data Visualization (Autovis): Driverless AI automatically selects data plots based on the most relevant data statistics, generates visualizations, and creates data plots that are most relevant from a statistical perspective based on the most relevant data statistics.

Automatic Feature Engineering: Software employs a library of algorithms and feature transformations to automatically engineer new, high value features for a given dataset.  Included in the interface is an easy-to-read variable importance chart that shows the significance of original and newly engineered features.

Automatic Model Documentation: Driverless AI provides an Autoreport (Autodoc) for each experiment, relieving the user from the time-consuming task of documenting and summarizing their workflow used when building machine learning models. The Autoreport includes details about the data used, the validation schema selected, model and feature tuning, and the final model created. With this capability in Driverless AI, practitioners can focus more on drawing actionable insights from the models and save weeks or even months in development, validation, and deployment process.

Time Series Forecasting:  Platform delivers superior time series capabilities to optimize for almost any prediction time window. Driverless AI incorporates data from numerous predictors, handles structured character data and high-cardinality categorical variables, and handles gaps in time series data and other missing values.

NLP with TensorFlow: Driverless AI automatically converts short text strings into features using powerful techniques like TFIDF. With TensorFlow, Driverless AI can also process larger text blocks and build models using all available data to solve business problems like sentiment analysis, document classification, and content tagging.

Automatic Scoring Pipelines: For completed experiments, it automatically generates both Python scoring pipelines and new ultra-low latency automatic scoring pipelines. The new automatic scoring pipeline is a unique technology that deploys all feature engineering and the winning machine learning model in a highly optimized, low-latency, production-ready Java code that can be deployed anywhere.

Machine Learning Interpretability (MLI): Driverless AI provides robust interpretability of machine learning models to explain modeling results in a human-readable format. In the MLI view, Driverless AI employs a host of different techniques and methodologies for interpreting and explaining the results of its models. A number of charts are generated automatically (depending on experiment type), including K-LIME, Shapley, Variable Importance, Decision Tree Surrogate, Partial Dependence, Individual Conditional Expectation, Sensitivity Analysis, NLP Tokens, NLP LOCO, and more.

Automatic Reason Codes: Reason codes show the key positive and negative factors in a model’s scoring decision in a simple language. Reasons codes are also useful in other industries, such as healthcare, because they can provide insights into model decisions that can drive additional testing or investigation.

Supported Algorithms

Constant Model  – predicts the same constant value for any input data. The constant value is computed by optimizing the given scorer. A constant model is meant as a baseline reference model. If it ends up being used in the final pipeline, a warning will be issued because that would indicate a problem in the dataset or target column (e.g., when trying to predict a random outcome).

Decision Tree – single (binary) tree model that splits the training data population into sub-groups (leaf nodes) with similar outcomes. No row or column sampling is performed, and the tree depth and method of growth (depth-wise or loss-guided) is controlled by hyper-parameters.

FTRL –  DataTable implementation of the FTRL-Proximal online learning algorithm. This implementation uses a hashing trick and Hogwild approach for parallelization. FTRL supports binomial and multinomial classification for categorical targets, as well as regression for continuous targets.

GLM – estimate regression models for outcomes following exponential distributions. GLMs are an extension of traditional linear models.

Isolation Forest – useful for identifying anomalies or outliers in data. Isolation Forest isolates observations by randomly selecting a feature and then randomly selecting a split value between the maximum and minimum values of that selected feature. This split depends on how long it takes to separate the points. Random partitioning produces noticeably shorter paths for anomalies. When a forest of random trees collectively produces shorter path lengths for particular samples, they are highly likely to be anomalies.

LightGBM – gradient boosting framework developed by Microsoft that uses tree based learning algorithms. It was specifically designed for lower memory usage and faster training speed and higher efficiency. Similar to XGBoost, it is one of the best gradient boosting implementations available. It is also used for fitting Random Forest, DART (experimental), and Decision Tree models inside of Driverless AI.

RuleFit –  creates an optimal set of decision rules by first fitting a tree model, and then fitting a Lasso (L1-regularized) GLM model to create a linear model consisting of the most important tree leaves (rules).

TensorFlow – open source software library for performing high performance numerical computation. Driverless AI includes a TensorFlow NLP recipe based on CNN Deeplearning models.

XGBoost -supervised learning algorithm that implements a process called boosting to yield accurate models. Boosting refers to the ensemble learning technique of building many models sequentially, with each new model attempting to correct for the deficiencies in the previous model. In tree boosting, each new model that is added to the ensemble is a decision tree. XGBoost provides parallel tree boosting (also known as GBDT, GBM) that solves many data science problems in a fast and accurate way. For many problems, XGBoost is one of the best gradient boosting machine (GBM) frameworks today. Driverless AI supports XGBoost GBM and XGBoost DART (experimental) models.

MLI for H2O Driverless AI

Driverless AI provides robust interpretability of machine learning models to explain modeling results in a human-readable format. In the Machine Learning Interpetability (MLI) view, Driverless AI employs a host of different techniques and methodologies for interpreting and explaining the results of its models. A number of charts are generated automatically (depending on experiment type), including K-LIME, Shapley, Variable Importance, Decision Tree Surrogate, Partial Dependence, Individual Conditional Expectation, Sensitivity Analysis, NLP Tokens, NLP LOCO, and more.

In Machine Learning Interpetability (MLI) view, #DriverlessAI generated automatically charts such as K-LIME, Shapley, Variable Importance, Decision Tree Surrogate, Partial Dependence, Individual Conditional Expectation, Sensitivity…
Click To Tweet

Driverless AI Scoring Pipelines

Driverless AI provides several Scoring Pipelines for experiments and/or interpreted models. A standalone Python Scoring Pipeline is available for experiments and interpreted models. The Python Scoring Pipeline is implemented as a Python whl file. While this allows for a single process scoring engine, the scoring service is generally implemented as a client/server architecture and supports interfaces for TCP and HTTP.

A low-latency, standalone MOJO Scoring Pipeline is available for experiments, with both Java and C++ backends.  The MOJO Scoring Pipeline provides a standalone scoring pipeline that converts experiments to MOJOs, which can be scored in real time. The MOJO Scoring Pipeline is available as either a Java runtime or a C++ runtime. For the C++ runtime, both Python and R wrappers are provided.

H2O AI customizations

The new bring-your-own-recipe capability ensures data scientists now can quickly customize and extend the Driverless AI to make their own AI with customizations – models, transformers, and scorers, extending the platform to meet any data science requirement. These customized recipes are then treated as first-class citizens in the automatic feature engineering process and eventually creating the winning model. With recipes domain experts now have more options to solve their data science problems and to make their own AI by enabling them to address a variety of use cases ranging from credit risk scoring, customer churn prediction, fraud detection, cyber threat prevention, sentiment analysis and more.

H2O also has the availability to set of vertical specific solutions: anti-money laundering, customer 360 & malicious domain detection. In addition, customers can explore and consume over 100 open-source recipes, curated by Kaggle Grandmasters at H2O.ai.

Project Workspace  for Model Operations, Administration, Collaboration

Project Workspace in Driverless AI enables data scientists to collaborate on different projects, build models, tag and version them appropriately for DevOps and IT that can then deploy these models to different environments in a highly scalable and robust manner.

In addition, H2O has a module for Model Admin that enables models that are deployed to be monitored for system health checks and also data science metrics around drift detection, model degradation, A/B testing, and provide alerts for recalibration and retraining.

Explainable AI

H2O Driverless AI provides robust interpretability of machine learning models to explain modeling results. H2O Driverless AI now has added the ability to perform disparate impact analysis to test for sociological biases in models.

This new feature allows for users to analyze whether a model produces adverse outcomes for different demographic groups even if those features were not included in the original model. These checks are critical for regulated industries where demographic biases in the data can creep into models causing adverse effects on protected groups.

#H2ODriverlessAI provides interpretable-by-design models including linear models, monotonic gradient boosting and RuleFit.
Click To Tweet

H2O Driverless AI provides interpretable-by-design models including linear models, monotonic gradient boosting and RuleFit. In its machine learning interpretability module, Driverless AI employs a host of different techniques and methodologies for explaining the results of its models like K-LIME, Shapley, variable importance, decision tree and partial dependence views.

Hardware and Driverless AI

Driverless AI runs on commodity hardware. It was also specifically designed to take advantage of graphical processing units (GPUs), including multi-GPU workstations and servers such as IBM’s Power9-GPU AC922 server and the NVIDIA DGX-1 for order-of-magnitude faster training.

BlueData

H2O.ai has an established partnership and several joint customers with BlueData.  With the BlueData EPIC software platform, data science teams can instantly spin up Driverless AI running on containers—whether on-premises, in the public cloud, or in a hybrid model.

HPE ProLiant DL380 Gen10 Server

The industry-leading server for multi-workload compute, the secure, resilient 2P/2U HPE ProLiant DL380 Gen10 Server features up to three NVIDIA Tesla GPUs and delivers world-class performance and supreme versatility for running AI/ML workloads such as H2O Driverless AI.

HPE Apollo 6500 Gen10 System

This system is a 4U dual-socket server featuring up to eight NVIDIA Tesla GPUs is an ideal high-performance computing (HPC) and deep learning platform. HPE Apollo 6500 Gen10 System provides unprecedented performance with industry-leading NVIDIA GPUs, fast GPU interconnect, high-bandwidth fabric, and a configurable GPU topology to match your workloads.

Installing and Upgrading Driverless AI

For the best  Driverless AI should be  installed on modern data center hardware with GPUs and CUDA support.  To simplify cloud installation, Driverless AI is provided as an AMI. To simplify local installation, Driverless AI is provided as a Docker image.

For native installs (rpm, deb, tar.sh), Driverless AI requires a minimum of 5 GB of system memory in order to start experiments and a minimum of 5 GB of disk space in order to run a small experiment. It is  recommended to have lots of system CPU memory (64 GB or more) and 1 TB of free disk space available. For Docker installs, it is  recommended 1TB of free disk space.

For running Driverless AI with GPUs, GPU should have compute capacity of at least 3.5 and at least 4GB of RAM. If these requirements are not met, then Driverless AI will switch to CPU-only mode.

Driverless AI supports local, LDAP, and PAM authentication. Authentication can be con_gured by setting environment variables or via a con_g.toml _le. Refer to the Setting Environment Variables section in the User Guide. Driverless AI also supports HDFS, S3, Google Cloud Storage, Google Big Query, KDB, Minio, and Snowake access.

About H2O.ai

H2O.ai is the open source leader in AI and automatic machine learning with a mission to democratize AI for everyone. H2O.ai is transforming the use of AI to empower every company to be an AI company in financial services, insurance, healthcare, telco, retail, pharmaceutical and marketing. H2O.ai is driving an open AI movement with H2O, which is used by more than 18,000 companies and hundreds of thousands of data scientists. H2O Driverless AI, an award winning and industry leading automatic machine learning platform for the enterprise, is helping data scientists across the world in every industry be more productive and deploy models in a faster, easier and cheaper way. H2O.ai partners with leading technology companies such as NVIDIA, IBM, AWS, Intel, Microsoft Azure and Google Cloud Platform and is proud of its growing customer base which includes Capital One, Nationwide, Walgreens and MarketAxess. H2O.ai believes in AI4Good with support for wildlife conservation and AI for academics.

Read more about DataRobot machine learning automation here

And here you can find information about open-source software for automated machine learning

Subscribe to Twitter to get latest updates:

Tweets by sakalosch

The post H2O Driverless AI For Machine Learning Automation appeared first on Quantum: Machine Learning & Analytics.

DataRobot automates the entire modeling lifecycle, enabling users to quickly and easily build highly accurate predictive models. DataRobot enables users to build and deploy highly accurate machine learning models in a fraction of the time it takes using traditional data science methods.

The DataRobot platform automatically searches through millions of combinations of algorithms, data preprocessing steps, transformations, features, and tuning parameters for the best machine learning model for the data.

The intuitive web-based interface allows anyone to interact with a very powerful platform, regardless of skill-level and machine learning experience. Users can drag-and-drop then let DataRobot do all the work or they can write their own models for evaluation by the platform.

DataRobot features a massively parallel modelling engine that can scale to hundreds or even thousands of powerful servers to explore, build and tune machine learning models.

Automated machine learning process

Machine Learning model can be deployed in 5 steps:

1. Uploading the dataset to DataRobot platform, which accepts input from a file, a remote URL, a JDBC data source or HDFS.
2. Platform infers the schema by suggesting appropriate data types for each feature. Business analysts and data scientists can perform necessary to advanced data exploratory activities on the ingested dataset. Finally, they need to select the target label which is going be predicted by the model.
3. After pressing Start, DataRobot creates a long queue of algorithms and trains each of them with the uploaded dataset.
4. Users can explore each model to understand the methodology and the parameters used in the training process. They can test each model with a test dataset to measure the accuracy.
5. Once a model has been chosen from the leaderboard, it is deployed as an API. The endpoint becomes ready to deal with production data. Any developer with the API key can invoke this like any other RESTful service.

DataRobot Features

Innovative open source algorithms. DataRobot uses the latest and most powerful open source machine learning libraries, including scikit-learn, H2O, TensorFlow, Vowpal Wabbit, Spark ML and XGBoost.

Automated feature engineering. DataRobot prepares data automatically, performing operations like one-hot encoding, missing value imputation, text mining, and standardization to transform features for optimal results.

Time-aware forecasting. DataRobot can automate the development of sophisticated time series models that predict the future values of a data series based on its history and trend. The platform automatically detects stationarity and seasonality, and implements backtesting to achieve the highest possible accuracy.

Multiclass model support. DataRobot allows for classification on targets with up to 100 distinct values, offering real-time and batch support for uncovering the predictive class and showing its probability across all classes.

Built-in guardrails. With DataRobot, modeling projects follow a consistent methodology based on data science  best practices. Novice users can’t “forget” to perform a critical step, such as model validation.

Advanced machine learning techniques. DataRobot incorporates the techniques advanced data scientists use: boosting, bagging, random forests, kernel-based methods, generalized linear models, deep learning, and many others.

Unsupervised anomaly detection. Anomalies can be uncovered in a dataset with DataRobot’s unsupervised ensemble blend model, which can offer new insights, even in familiar datasets.

Manual tuning capabilities. DataRobot automates model tuning, but also supports manual tuning so you can tune and adjust machine learning algorithms for even better results.

Visual AI

Visual AI provides the ability to include images in your Supervised Machine Learning pipeline. Similar to other DataRobot projects, Visual AI projects deliver both deployable models and associated model insights.

Visual AI has a wide variety of highly efficient, state-of-the-art, deep learning featurizers including Squeezenet, resnet50, xception, efficient-net, and some others. Each of these featurizers convert an input image to a vector of numbers

Visual AI has extra tools that are specific to the image data type which were created to enhance model insights.

• Image Activation Maps allow you to see sample locations in the image that the model is using to make decisions.
• Image Embeddings allow you to visualize a sample of images projected from their original N-dimensional feature space to a new 2-dimensional feature space. This makes it easy to see what images are considered similar.

Enterprise application

Deployment of DataRobot on-premise on standalone servers, an existing Hadoop infrastructure, or in a Virtual Private Cloud (VPC).  It is also available as a managed SaaS offering hosted on Amazon Web Services (AWS).

DataRobot can be installed as a service on YARN in Hadoop clusters, and perform distributed model scoring on data stored on HDFS.

DataRobot offers fine-grained role-based security, including two-factor authentication, and supports Kerberos and LDAP protocols.

DataRobot’s MLOps

DataRobot’s MLOps capabilities are integrated into the DataRobot Enterprise AI Platform and allow to manage all production models from one place. They help with flexible deployment of custom models built with Python, R or other compatible machine learning platforms.Centralized management of models, embedded anywhere, built by anyone can be carried out across organization.

MLOps also ensure proactive management of  production models to prevent production issues and ensuring both model trust and performance. Real-time dashboards with automated monitoring and alerts provide information on data deviations and key model metrics to quickly and proactively adapt to changing conditions. This also provide deep production diagnostics in order to improve failure prediction, minimize SLA violations, and optimize system operations.

In the meantime, MLOps guarantee safe scaling of AI projects and maintain control over production models and compliance with regulations. Access to production models and systems could be appropriately limited  to manage organizational risk and satisfy regulatory requirements. Individual predictions is possible to trace back to the production model used to make each prediction to ensure legal and regulatory compliance. MLOps also keep an audit trail over the the lifetime of a model deployment showing when and where the model was deployed, who made updates and why.

Paxata for automated data preparation

Paxata is the software that conducts automatically raw data preparation which include finding, ingesting, profiling, cleaning, and transforming raw data with an enterprise-grade, self-service data preparation application. Paxata Adaptive Information Platform can create actionable information in real-time with the help machine learning, an intuitive user experience, and a smart distributed architecture.

It can ingest data from a wide variety of enterprise sources, including complex semi-structured files such as XML or JSON,NoSQL and relational databases, and cloud applications. Paxata intelligently detects data source types and transforms it into a tabular format for point-and-click interaction and profiling.

Paxata profiles data and generates a scorecard showing data type distribution, field completeness, field length analysis, top/ bottom patterns, leading or trailing string patterns, min/max and range, special character analysis, custom calculations, and more. Its interactive, Excel-like interface allows you to search, investigate and discover trends, outliers, and patterns across entire data set – not just a sample, and validate your data visually for immediate feedback. You can also combine and blend multiple, mixed-type data sources together.

The software standardize similar values and misspellings, joins, appends, and overlaps across data sources with smart machine learning recommendations. Then it-documents your steps to create repeatability, auditing, and governance.

Exporting and publishing of result datasets is possible to a broadrange of databases, applications, AI platforms like DataRobot or popular BI tools, such as Tableau, Qlik, MicroStrategy, PowerBI, and many others.

DataRobot Integration with  Factset

DataRobot on FactSet integrates automated machine learning technology into the FactSet workstation, which allows you to build, deploy, monitor, and manage sophisticated machine learning models quickly and easily. Access hundreds of powerful open source machine learning algorithms to create predictive, automated AI applications for factors such as equity volatility, bond performance, and macroeconomic event predictions. This allows to:

Create AI-driven investment decision workflows. Combination of DataRobot’s automated machine learning with FactSet’s content and applications to embed AI directly into data exploration, signal generation, and evaluation workflows in an integrated manner.

Connect core datasets to advanced machine learning algorithms. Power automated machine learning workflow with core datasets including fundamentals and estimates, premium unique content such as sentiment and ESG, and your own proprietary content.

Enhance factor research with automated machine learning. Easily build and deploy machine learning models to produce more intelligent alpha signals, risk forecasts, macroindicators within FactSet applications spanning the entire portfolio lifecycle.

About DataRobot

DataRobot is a Boston-based tech company with offices in New York, London, Kyiv, Singapore, Tokyo, and Sydney. Backed by prominent venture investors like NEA, Atlas Ventures, and TechStars, it developes an automated machine learning platform that helps data scientists and analysts of all levels build and deploy better predictive models and improve discoverability of valuable business insights. DataRobots works with global brands, like Mitsubishi Heavy Industries, Nippon Steel, Airbnb, United Airlines, Panasonic, and others. It enable advanced predictive analytics at many Fortune 500 enterprises, including largest US banks, insurers, fintechs, energy companies, and healthcare institutions.

You can read more about automated machine learning here

Subscribe to Twitter to get latest updates:

Tweets by sakalosch

The post DataRobot And Paxata For Automation Of Machine Learning And Data Preparation appeared first on Quantum: Machine Learning & Analytics.

Forge allows enterprise users with no presumed quantum computing expertise to run problems on a wide range of quantum computing hardware platforms and simulators. Forge users can access end-to-end implementations of proprietary and open-source algorithms for binary optimization, chemistry simulation, and machine learning.Forge has three main classes of algorithms including binary optimization, chemistry simulation, and machine learning.

Forge provides two key features that distinguish it from current market offerings:

• enables users to call high-level algorithms on top of what quantum computing hardware vendors offer in their software development kits
• integrates a wide variety of quantum computing hardware chips and software simulators. Users can easily target these backends to run their problems

Through Forge, customers can gain access to a library of the quantum algorithms that QC Ware has spent the past years developing in collaboration with its customers. For QC Ware’s customers, those algorithms can be integrated to suit their needs, and once it’s ready, Forge can run them either on a high-performance computer capable of simulating quantum operations or on an actual quantum computer. Customers will be able to choose to run on machines owned by D-Wave, IBM or Rigetti.  

QC Ware is a quantum-computing-as-a-service company based in Palo Alto, California. QC Ware develops hardware-agnostic enterprise software solutions for quantum computers. The company’s quantum computing software solves problems in combinatorial optimization and machine learning with efficiency unmatched by traditional HPC solutions.

The post FORGE -Data Science Platform For Quantum Computing appeared first on Quantum: Machine Learning & Analytics.

It enables users to get more value out of data, including OSIsoft PI System, Honeywell’s UniformancePHD, Emerson DeltaV and Ovation, Inductive Automation’s Ignition, AspenTech IP.21, Wonderware, GE Proficy or any other, by providing interactive, visual tools to accelerate industrial process analytics.

Seeq can be set up and running on a dedicated server, server cluster, or virtual machine in as little time as an hour. On-premise installations on the same network as a plant or enterprise historian, or on the cloud (Microsoft Azure, Amazon Web Services, etc.), or on a mixed environment of on-premise and cloud resources are supported.

This software is extensible through data export, data integration, and a REST API for creating custom templates and modules. Data export options include Microsoft Excel and PowerPoint, and any OData client (Tableau, Microsoft PowerBI, etc.). Data integration with OSIsoft Vision is supported, and the REST API has SDKs for programming in C#, Python, MatLab, and Java.

Seeq does not copy or duplicate data from the source of record. Instead, data is accessed via a connector retrieving data on the fly based on user actions. All Seeq documents such as workbooks, worksheets, topics, and search definitions are stored by Seeq for easy user access and sharing.

The post Industrial Data Analytics With Powerful SEEQ Software appeared first on Quantum: Machine Learning & Analytics.

Kendra understands natural language questions to get the answers, whether that is a precise answer, an FAQ, or an entire document. It lets easily add content from file systems, SharePoint, intranet sites, file sharing services, and more, into a centralized location in order to allow quickly search all of the information to find the best answer.

Results get better over time search, because Kendra’s machine learning algorithms learn which results your users find most valuable. Kendra actively retrains deep learning models built for data set and employee usage patterns to improve search accuracy. As end-users interact with search results, Kendra fine tunes its results. Kendra gives the option to manually tune relevance; it is possible to boost certain fields in index like document freshness, view counts, or specific data sources.

Kendra helps ensure that search results adhere to existing access policies by scanning permissions on documents so that results only contain documents the user has permission to access, and it encrypts data in transit and at rest.

Kendra’s models are optimized to understand language from domains like IT, health care, and insurance, plus energy, industrial, financial services, legal, media and entertainment, travel and hospitality, human resources, news, telecommunications, mining, food and beverage, and automotive. Beyond Kendra, Microsoft has launched Project Cortex, a service that taps AI to automatically classify and analyze an organization’s documents, conversations, meetings, and videos. It was in some ways a direct response to Google Cloud Search, which pulls in data from a range of third-party products and services running both on-premises and in the cloud, relying on machine learning to deliver query suggestions and surface the most relevant results.

Twitter

LinkedIn

Facebook

The post Amazon launched Kendra, machine learning-powered service for enterprise search appeared first on Quantum: Machine Learning & Analytics.

• The global context-aware processing market will reach $7.15B by 2025
• AI in IoT Devices will exceed $105B in North America by 2025 with 28% CAGR
• The AI in manufacturing market will reach $20.2B globally by 2025 with 35.3% CAGR
• Asia-Pac and North America AI markets will reach $59B and $55.4B respectively by 2025
• Starting with consumer products AIoT will be a significant driver of AI adoption in enterprise, industrial, and government

AI Chipsets: Multi-processor AI chipsets learn from the environment, users, and machines to uncover hidden pattern among data, predict actionable insight, and perform actions based on specific situations. AI chipsets will become an integral part of both AI software/systems as well as critical support of any data-intensive operation as they drastically improve processing for various functions as well as enhance overall computing performance.

AI in Edge Networks: Edge networks will be the point of the spear so to speak when it comes to data handling, meaning that streaming data will be available for processing and decision-making. While advanced data analytics software solutions can be very effective for this purpose, there will be opportunities to enhance real-time data analytics by way of leveraging AI to automate decision making and to engage machine learning for ongoing efficiency and effectiveness improvements.

AI and 5G: The role and importance of AI in 5G ranges from optimizing resource allocation to data security and protection of network and enterprise assets.

AI and Real-time Data Processing: The ability to capture streaming data, determine valuable attributes, and make decisions in real-time will add an entirely new dimension to service logic. In addition, AI will support data management across all of these areas. The growing amount of human-oriented and machine-generated from communications, applications, content, and commerce data will drive substantial opportunities for AI support of unstructured data analytics solutions.

The outlook for AI in support of the ICT industry is strong, especially when one considers that the purpose of telecom and IT services is to support virtually every other industry in terms of communications, applications, content, and commerce.

The post AI Industry appeared first on Quantum: Machine Learning & Analytics.

WaveNetEQ is a generative model, based on DeepMind’s WaveRNN technology, that is trained using a large corpus of speech data to realistically continue short speech segments enabling it to fully synthesize the raw waveform of missing speech. Because Duo calls are end-to-end encrypted, all processing needs to be done on-device. The WaveNetEQ model is fast enough to run on a phone, while still providing state-of-the-art audio quality and more natural sounding PLC than other systems currently in use.

WaveNetEQ uses the autoregressive network to provide the audio continuation during a packet loss event and the conditioning network to model long term features, like voice characteristics. The spectrogram of the past audio signal is used as input for the conditioning network, which extracts limited information about the prosody and textual content. This condensed information is fed to the autoregressive network, which combines it with the audio of the recent past to predict the next sample in the waveform domain. This differs slightly from the procedure that was followed during training of the WaveNetEQ model, where the autoregressive network receives the actual sample present in the training data as input for the next step, rather than using the last sample it produced. This process, called teacher forcing, assures that the model learns valuable information, even at an early stage of training when its predictions are still of low quality. Once the model is fully trained and put to use in an audio or video call, teacher forcing is only used to “warm up” the model for the first sample, and after that its own output is passed back as input for the next step.

The model is applied to the audio data in Duo’s jitter buffer. Once the real audio continues after a packet loss event, we seamlessly merge the synthetic and real audio stream. In order to find the best alignment between the two signals, the model generates slightly more output than is required and then cross-fades from one to the other. This makes the transition smooth and avoids noticeable noise.

The post WaveNetEQ appeared first on Quantum: Machine Learning & Analytics.