Topics

Deals

100 tips for ML

Machine Learning Formal definition

A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P if its performance at tasks in T, as measured by P, improves with experience E.�

-------------

Machine Learning definition

Computer finds pattern in data and using pattern it learn and predict future outcome

-------------

Machine Learning definition

-------------

Approaches in Machine Learning

  • Supervised Learning
  • Un Supervised Learning
  • Supervised Learning

    In supervised learning there is labelled data. Algorithm learn from past example and preict future. Example - Hand writing recognition

    In un supervised learning there is no labelled data. Algorithm learns, understand and make sense of data and provide insight - example -Anomaly detection

    -------------

    Machine Learning capabilities

  • Classification
  • Regression
  • Clustering
  • Anomaly
  • Which Action
  • -------------

    Steps to build ML experiment

  • Define the Problem
  • Create a Hypothesis
  • Create a experiment that you can repeat and have control
  • Determine result
  • Iterate
  • -------------

    Famous Machine Learning conferences

  • Data Science Strata
  • ICLR Deep learning
  • Seattle Data Day
  • KDD
  • AISTATS
  • ICML/NIPS
  • UAI
  • COLT/ALT
  • DMBI - Annually
  • -------------

    What are skills needed in ML/Data science expert

  • Data
  • Math
  • Problem solving
  • Domain expertise
  • Data

  • Learn method to process structured, unstructured, semi structured data
  • Know Big Data processing skills and architecture
  • modelling technqiues, database design, No-SQL
  • Machine Learning
  • Languages like R, Python or Machine Learning tools
  • Math

  • statistics
  • Linear Algebra
  • Logistics and Sets
  • Problem solving

  • Passion to solve problem
  • Domain expertise

    There is need and demand to know specific industry

  • Finance
  • Manufacturing
  • Professional services
  • Business Process and Scenarios

    One can have specialization in a business process or scenario such as :

  • Fraud detection
  • Lead generation
  • Sales Pipeline
  • Predictive maintenance
  • Revenue forecasting
  • Product recommendation
  • -------------

    Regression vs Machine Learning

    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------
    -------------

    Microsoft R Server

    Microsoft R Server is one of solution to build R solution for large datasets.

    -------------

    5 Tribes of Machine Learning

    -------------
    -------------



    Like us on facebook


    Facebook comments

    Articles in