Platform of Machine Learning

1
UNIT -II
2

• Platform for machine learning -Machine learning
  python libraries - Scikit-learn -training data
  testing data validation data - k-fold cross
  validation -Features -Performance metrics -
  MSE, accuracy, confusion matrix, precision -
  recall, F- score - Linear Regression with
  multiple variables - Logistic Regression - spam
  filtering with logistic regression-Naive Bayes
  with scikit-learn

3
PLATFORM FOR MACHINE LEARNING

• Best Machine Learning Platforms
• KNIME Analytics Platform
• KNIME Analytics Platform is a free open-source
  platform that provides end-to-end data analysis,
  integration, and reporting. With the KNIME
  Analytics Platform, data scientists can easily
  enable the creation of visual workflows via a
  drag-and-drop-style graphical interface.
• It will not require knowledge of coding.
• To build workflows, a user gets to choose from
  more than 2000 nodes. KNIME Analytics allows
  developers to perform various actions from basic
  I/O to data manipulations, transformations, and
  data mining.
• The best part of KNIME Analytics is that it
  consolidates the entire function process into a
  single workflow.

4

• Top Features
• Parallel execution on multi-core systems
• Offers scalability through sophisticated data
  handling
• Simple extensibility via a well-defined API for
  plugin extensions

5
TIBCO Software

• TIBCO is a data science platform that supports
  the entire analytics lifecycle with capabilities
  to include cloud-based analytics that integrates
  with many open source libraries.
• TIBCO data science allows the user to prepare
  data and do the model building, deployment, and
  monitoring. It is widely known for use cases,
  such as product refinement and business
  exploration.
• Top Features
• Automatically detects the locations and
  visualizes those data as an interactive map
• Can analyze data in many visualization types,
  such as tables and charts
• Can access real-time streaming and spot issues

6
Amazon SageMaker

• Amazon SageMaker is a cloud machine-learning
  platform that allows developers to create, train,
  and deploy machine learning models. Data
  scientists or developers can easily deploy
  machine learning models on embedded systems and
  edge devices.
• It is developed by Amazon Web Services (AWS) that
  offers the broadest machine and deepest set of
  learning services and supporting cloud
  infrastructure.
• Top Features
• Build highly accurate training datasets
• Provides a visual interface to analyze the debug
  data for the easy inspection process
• No need to break down the data set into multiple
  chunks

7
Alteryx Analytics

• Alteryx is an analytic platform that accelerates
  digital transformation. It offers data
  accessibility and data science processes.
• Alteryx provides data scientists with a platform
  that helps them build models in a workflow.
• Their vision is to make it easy to use for
  companies to cultivate a data analytics culture
  without the requirement of data scientists. In
  self-service data analytics, Alteryx always
  leads.
• Top Features
• Automate manual data tasks into repeatable
  analytics workflows
• Freedom to deploy and manage analytic models
• Support almost every data source and
  visualization tool

8
SAS

• SAS is a data science and analytics software
  vendor that offers a robust suite of advanced
  analytics and data science products. The best
  part of considering the SAS platform is that you
  can easily access data in any format and from any
  source.
• It automatically generates a pipeline that
  dynamically adjusts to the data. It also includes
  natural language generation for project
  management. SAS Model Management allows users to
  register SAS and open-source models within
  projects or as standalone models.
• Top Features
• Allows to explore data from within Model Studio
  and launch directly into SAS Visual Analytics
• Visual interface for the entire analytical
  lifecycle process
• View training data within each node in Model
  Studio

9
H2O.ai

• H2O.ai offers a wide range of AI platforms and
  data science platforms to expert data scientists.
  It is a fully open-source platform, which is a
  distributed in-memory ML platform with linear
  scalability.
• H2O supports statistical and machine learning
  algorithms for those who are looking for deep
  learning platforms because deep learning is one
  of the significant advantages of it.
• It offers Deep Water for deep learning, where H2O
  continues to lead in data science and machine
  learning unified platforms.
• Top Features
• Process large text blocks and build models with
  Natural Language Processing
• Has image processing capabilities using over 30
  pre-trained image transformers and models
• Automatically generates visualizations and
  creates data plots

10
Databricks Unified Analytics Platform

• It offers a cloud and Apache Spark-based unified
  analytics platform, which is a combination of
  data engineering and data science functionality.
• The one who is working on the Databricks Unified
  Analytics platform gets exclusive features for
  real-time enablement on AWS, high performance,
  and best operation results.
• One can simply access the preconfigured ML
  environments with a single-click feature for
  Augmented Machine Learning with popular
  frameworks.
• Top Features
• Highly reliable and performant data pipelines
• Productive data science at scale
• End-to-end data security and compliance

11
Microsoft Azure Machine Learning Studio

• Microsoft offers its services in the data science
  and machine-learning markets through its Azure
  software products.
• The cloud-based Azure Machine Learning Studio is
  perfectly designed for those who want to build,
  test, and execute predictive analytics, and
  machine learning solutions on their data. Azure
  is one of the most liked platforms in the data
  science and machine learning markets.
• Apart from this, Microsoft Azure allows customers
  to easily build, deploy, and share advanced
  analytics solutions in the cloud.
• Top Features
• Integrates with existing Develop processes and
  helps to manage the complete ML lifecycle
• Understand models with interpretability and
  protect data
• Provides the best support for open-source
  frameworks and languages

12
DataRobot

• DataRobot offers an enterprise AI machine
  learning platform that makes it fast and easy to
  maintain, build, and deploy AI.
• It includes Automated Machine Learning, Automated
  Time Series, and ML Ops which are independent and
  fully integrated tools.
• All can be easily deployed in various ways
  according to the business requirements. It helps
  you to solve business problems by finding the
  best predictive model for your data.
• Top Features
• Allows to build innovative new models from
  incredibly diverse types of data
• Monitor models via a centralized dashboard to see
  usage in real-time
• Provides fully explainable AI through
  human-friendly visual insights

13
Rapid Miner

• RapidMiner offers a data science platform to help
  people for building and operating AI solutions.
  It covers the entire lifecycle of the AI
  production process from data exploration and
  data preparation to model building, ML models
  deployment, and model operations.
• It comes with the depth that is essential for a
  data scientist to simplify AI via a visual user
  interface that can easily streamline the process
  of building and understanding complex models.
• Top Features
• The powerful visual programming environment
• Access, load, and analyze any type of data
• Efficiently build and deliver better models
  faster