Enhanced Machine Learning

Description:

This intensive one-day workshop is crafted to deepen participants' understanding of advanced machine learning concepts and techniques. The class will cover crucial topics including handling imbalanced datasets, hyperparameter tuning, dimensionality reduction, TensorFlow tensors, understanding neural networks, and activation functions. Through a blend of theoretical insights and practical exercises, attendees will gain hands-on experience and actionable knowledge to enhance their machine learning proficiency.

NOTE: This course is not meant to cover Generative AI and is not intended to cover unstructured data such as images and text.

Duration: 1 Day

Course Code: BDT364

Learning Objectives:

After this course, you will be able to:

• Handling Imbalanced Datasets
• Implementing Hyperparameter Tuning
• Applying Dimensionality Reduction
• Working with TensorFlow Tensors
• Comprehend Neural Networks, explore activation functions
• Build and Train Neural Networks

• Basic understanding of machine learning concepts (or completion of the “Kickstart AI and Machine Learning” course). No prior knowledge of neural networks is required.

• This course is designed for Data Scientists, Data Engineers, Software Engineers, Software Architects, and Quality Assurance Engineers who have completed the “Kickstart AI and Machine Learning” course.

• This course is designed for Data Scientists, Data Engineers, Software Engineers, Software Architects, and Quality Assurance Engineers who have completed the “Kickstart AI and Machine Learning” course.

Course Outline:

1. Handling Imbalanced Datasets

• Understanding challenges for classification problems due to imbalanced target variable
• Apply techniques to handle imbalanced datasets
• Lab: Visualization and applying different techniques to resolve this problem

2. Implementing Hyper-parameter tuning

• Learning importance of hyper parameters on model performance
• Use techniques for hyperparameter tuning such as Grid Search and K-Folds
• Lab: To apply Grid Search and K-Fold to various machine learning algorithms

3. Applying Dimensionality Reduction

• Understand dimensionality reduction on high dimensional data
• Implement common dimensionality reduction techniques
• Lab: Applying Principal Component Analysis (PCA)

4. Working with TensorFlow Tensors

• Gain practical experience with TensorFlow tensors, its creation, manipulating them
• Lab: Working with Tensors

5. Comprehend Neural Networks & Activation Functions

• Get an overview of neural network architecture and its components
• Understand how neural networks learns from data and their role in various machine learning tasks
• Understand various activation functions used in neural networks
• Lab: Activations functions and basic neural network

6. Build and Train Neural Networks

• Learn the process of designing and constructing neural network models
• Understand layer configurations, activation functions, optimizers and training procedures
• Lab: Implement a neural network using TensorFlow and Keras

Training material provided: Yes (Digital format)

Hands-on Lab: All labs will be conducted using Google Colaboratory. Participants must have a Google Email ID to access the lab environment.