Advanced Predictive Modelling in R Certification Training

Advanced Predictive Modelling in R Certification Training

R offers a free and open source environment that is perfect for both learning and deploying predictive modelling solutions. This Certification Training is intended for a broad audience as both, an introduction to predictive models as well as a guide to applying them, covering topics such as Ordinary Least Square Regression, Advanced Regression, Imputation, Dimensionality Reduction etc. Readers will also be able to learn basics of Statistics, such as Correlation and Linear Regression Analysis.

Basic Statistics in R

Learning Objectives: In this module you will get a brief introduction to statistics and will conduct best test and exploratory analysis.

Topics:

• Covariance & Correlation
• Central Limit Theorem
• Z Score
• Normal Distributions
• Hypothesis

Hands On: Calculating statistical parameters such as mean, median, mode and making custom visualizations for developing intuition of data with respect to statistical parameters.

Ordinary Least Square Regression 1

Learning Objectives: In this module you will get a brief introduction of basic regression and multiple regression and will learn how to present the same graphically.

Topics:

• Bivariate Data
• Quantifying Association
• The Best Line: Least Squares Method
• The Regressions
• Simple Linear Regression
• Deletion Diagnostics and Influential Observations
• Regularization

Hands On: Ridge and Lasso regression implementation.

Ordinary Least Square Regression 2

Learning Objectives: The goal of this module is to dive you into linear regression and make the model a better fit, make necessary transformation check for over fitting and under fitting and outliers’ identification and treatment.

Topics:

• Model fitting using Linear Regression
• Performing Over Fitting & Under Fitting
• Collinearity
• What is Heteroscedasticity?

Hands On: Perform exploratory data analysis and check for heteroscedasticity, perform remedial steps and transform the data and implement linear regression model.

Logistic Regression

Learning Objectives: In this module, you will understand the problems related with Linear Probability Model, will be introduced to logistic regression and various uses of the same and its industry usage.

Topics:

• Binary Response Regression Model
• Linear regression as Linear Probability Model
• Problems with Linear Probability Model
• Logistic Function
• Logistic Curve
• Goodness of fit matrix
• All Interactions Logistic Regression
• Multinomial Logit
• Interpretation
• Ordered Categorical Variable

Hands On: Build a logistic regression model to classify the data.

Advanced Regression

Learning Objectives: In this module, you will dig deeper into logistic regression and learn about more varied usage of logistic regression on various dataset.

Topics:

• Poisson Regression
• Model Fit Test
• Offset Regression
• Poisson Model with Offset
• Negative Binomial
• Dual Models
• Hurdle Models
• Zero-Inflated Poisson Models
• Variables used in the Analysis
• Poisson Regression Parameter Estimates
• Zero-Inflated Negative Binomial

Hands On: Create ZINB and Hurdel regression model.

Imputation

Learning Objectives: In this module, you will learn about addressing missing values and how to impute it using various processes.

Topics:

• Missing Values are Common
• Types of Missing Values
• Why is Missing Data a Problem?
• No Treatment Option: Complete Case Method
• No Treatment Option: Available Case Method
• Problems with Pairwise Deletion
• Mean Substitution Method
• Imputation
• Regression Substitution Method
• K-Nearest Neighbour Approach
• Maximum Likelihood Estimation
• EM Algorithm
• Single and Multiple Imputation
• Little’s Test for MCAR

Hands On: Implement KNN model and perform single and multiple imputation.

Forecasting 1

Learning Objectives: The goal of this module is to give an introduction on forecasting and time series data.

Topics:

• Need for Forecasting
• Types of Forecast
• Forecasting Steps
• Autocorrelation
• Correlogram
• Time Series Components
• Variations in Time Series
• Seasonality
• Forecast Error
• Mean Error (ME)
• MPE and MAPE—Unit free measure
• Additive v/s Multiplicative Seasonality
• Curve Fitting
• Simple Exponential Smoothing (SES)
• Decomposition with R
• Generating Forecasts
• Explicit Modeling
• Modeling of Trend
• Seasonal Components
• Smoothing Methods
• ARIMA Model-building

Hands On: Implement Exponential Smoothing and ARIMA model for time series forecasting.

Forecasting 2

Learning Objectives: In this module, you will learn about Seasonality, Trend Analysis and decaying the factors over the time.

Topics:

• Analysis of Log-transformed Data
• How to Formulate the Model
• Partial Regression Plot
• Normal Probability Plot
• Tests for Normality
• Box-Cox Transformation
• Box-Tidwell Transformation
• Growth Curves
• Logistic Regression: Binary
• Neural Network
• Network Architectures
• Neural Network Mathematics

Dimensionality Reduction

Learning Objectives: In this module, you will get a complete knowledge on Dimensionality Reduction and will discuss and apply few of the important algorithms associated with Dimensionality Reduction.

Topics:

• Factor Analysis
• Principal Component Analysis
• Mechanism of finding PCA
• Linear Discriminant Analysis (LDA)
• Determining the maximum separable line using LDA
• Implement Dimensionality Reduction algorithm in R

Hands On: Implement Principal component analysis and Boosting (ADAboost).

Survival Analysis

Learning Objectives: In this module, you will learn about Churn analysis and Regression on time series data with time component.

Topics:

• Time-to-Event Data
• Censoring
• Survival Analysis
• Types of Censoring
• Survival Analysis Techniques
• PreProcessing
• Elastic Net

Hands On: Do PCA preprocessing and implement Elastic Net model.