DATASCIENCE CONTENT:

 DESCRIPTIVE STATISTICS AND PROBABILITY DISTRIBUTIONS:

• Introduction about Statistics
• Different Types of Variables
• Measures of Central Tendency with examples
• Measures of Dispersion
• Probability & Distributions
• Probability Basics
• Binomial Distribution and its properties
• Poisson distribution and its properties
• Normal distribution and its properties

INFERENTIAL STATISTICS AND TESTING OF HYPOTHESIS

• Sampling methods
• Different methods of estimation
• Testing of Hypothesis & Tests
• Analysis of Variance

COVARIANCE & CORRELATION

PREDICTIVE MODELING STEPS AND METHODOLOGY WITH LIVE EXAMPLE:

• Data Preparation
• Exploratory Data analysis
• Model Development
• Model Validation
• Model Implementation

SUPERVISED TECHNIQUES:

MULTIPLE LINEAR REGRESSION

• Linear Regression - Introduction - Applications
• Assumptions of Linear Regression
• Building Linear Regression Model
• Understanding standard metrics (Variable significance, R-square/Adjusted R-Square, Global hypothesis etc)
• Validation of Linear Regression Models (Re running Vs. Scoring)
• Standard Business Outputs (Decile Analysis, Error distribution (histogram), Model equation, drivers etc)
• Interpretation of Results - Business Validation - Implementation on new data
• Real time case study of Manufacturing and Telecom Industry to estimate the future revenue using the models

LOGISTIC REGRESSION - INTRODUCTION - APPLICATIONS

• Linear Regression Vs. Logistic Regression Vs. Generalized Linear Models
• Building Logistic Regression Model
• Understanding standard model metrics (Concordance, Variable significance, Hosmer Lemeshov Test, Gini, KS, Misclassification etc)
• Validation of Logistic Regression Models (Re running Vs. Scoring)
• Standard Business Outputs (Decile Analysis, ROC Curve)
• Probability Cut-offs, Lift charts, Model equation, drivers etc)
• Interpretation of Results - Business Validation - Implementation on new data
• Real time case study to Predict the Churn customers in the Banking and Retail industry

PARTIAL LEAST SQUARE REGRESSION

• Partial Least square Regression - Introduction - Applications
• Difference between Linear Regression and Partial Least Square Regression
• Building PLS Model
• Understanding standard metrics (Variable significance, R-square/Adjusted R-Square, Global hypothesis etc)
• Interpretation of Results - Business Validation - Implementation on new data
• Sharing the real time example to identify the key factors which are driving the Revenue

VARIABLE REDUCTION TECHNIQUES

FACTOR ANALYSIS

PRINCIPLE COMPONENT ANALYSIS

• Assumptions of PCA
• Working Mechanism of PCA
• Types of Rotations
• Standardization
• Positives and Negatives of PCA

SUPERVISED TECHNIQUES CLASSIFICATION:

CHAID

CART

DIFFERENCE BETWEEN CHAID AND CART

RANDOM FOREST

• Decision tree vs. Random Forest
• Data Preparation
• Missing data imputation
• Outlier detection
• Handling imbalance data
• Random Record selection
• Random Forest R parameters
• Random Variable selection
• Optimal number of variables selection
• Calculating Out Of Bag (OOB) error rate
• Calculating Out of Bag Predictions

COUPLE OF REAL TIME USE CASES WHICH ARE RELATED TO TELECOM AND RETAIL INDUSTRY. IDENTIFICATION OF THE CHURN.

UNSUPERVISED TECHNIQUES:

SEGMENTATION FOR MARKETING ANALYSIS

•  Need for segmentation
• Criterion of segmentation
• Types of distances
• Clustering algorithms
• Hierarchical clustering
• K-means clustering
• Deciding number of clusters
• Case study

BUSINESS RULES CRITERIA

REAL TIME USE CASE TO IDENTIFY THE MOST VALUABLE REVENUE GENERATING CUSTOMERS.

TIME SERIES ANALYSIS:

TIME SERIES COMPONENTS( TREND, SEASONALITY, CYCLICITY AND LEVEL) AND DECOMPOSITION

BASIC TECHNIQUES

• Averages,
• Smoothening etc

ADVANCED TECHNIQUES

• AR Models,
• ARIMA
• UCM
• Hybrid Model

UNDERSTANDING FORECASTING ACCURACY - MAPE, MAD, MSE ETC

COUPLE OF USE CASES, TO FORECAST THE FUTURE SALES OF PRODUCTS

TEXT ANALYTICS:

GATHERING TEXT DATA FROM WEB AND OTHER SOURCES

PROCESSING RAW WEB DATA

COLLECTING TWITTER DATA WITH TWITTER API

NAIVE BAYES ALGORITHM

• Assumptions and of Naïve Bayes
• Processing of Text data
• Handling Standard and Text data
• Building Naïve Bayes Model
• Understanding standard model metrics
• Validation of the Models (Re running Vs. Scoring)

SENTIMENT ANALYSIS

• Goal Setting
• Text Preprocessing
• Parsing the content
• Text refinement
• Analysis and Scoring