Predictive Modeling and Machine Learning

Overview:

Introduction:

This course uses case studies, exercises, and interactive discussion. Each machine learning method has a case study to back it up, along with step-by-step outputs that run concurrently with its multi-stage analysis. On comparable technologies like SPSS, SAS, Statistica, and Excel, all algorithms are described with consecutive screen shots.

Course Objectives:

At the end of this course the participants will be able to: 

• Learn what machine learning really means.
• Understand the main distinctions between machine learning and data analysis.
• Integrate testing and sample validation into machine learning models
• Please provide a summary of the top analytical solutions.
• Apply precise estimation using comprehensive prediction models.

Targeted Audience:

This training would be useful for any level of professional interested in how machine learning may help their organization. These include experts from a variety of fields, such as finance, insurance, retail, government, manufacturing, healthcare, telecom, and aviation, among others.

Course Outline:

Unit1:Analysis of the Data and Simple Regression

• the means and proportions of two groups are examined
• Using a single chart to profile two groups
• comparing the averages and proportions of several groupings
• combining numerous group profiles into a single chart
• an easy regression
• Regression versus correlation
• Sensitivity evaluation of numerical variables

Unit2: Logistic and Multiple Regressions

• Machine Learning Introduction
• Logic of Gradient Descent
• Regression Comparison: Multiple vs. Simple
• estimate variability analysis
• fake variables
• Logistic and multiple regressions: Similarities and Disparities
• reducing complexity of models
• Regression with steps

Unit3: Analysis of Discrimination

• Enhanced Profiling
• Discriminant Function for Two Groups
• Assignment of Cases
• Model Assessment
• Functions of classification
• Squared Mahalanobis Distances
• probabilistic approach
• Reduction of the Model
• Common Discriminant Analysis

Unit4: Decision Trees

• How do decision trees work?
• Borland Trees
• Decision Tree's Qualities
• prudence guidelines
• Classification Tree (CART)
• Regression Tree, CART
• Tree CHAID
• Unexpected Forest Tree

Unit5: Bayesian, neural networks, nearest neighbor, and deep learning

• Conditions for probabilities
• Probabilistic forecasting
• Distance between neighbors
• K-nearest neighbors' distances from you
• Weights in a Neural Network model
• Hidden layers play a role
• Advantages and disadvantages of neural networks
• Deep Learning
• Introduction to Big Data