CSIT-602 - Machine Learning

Define machine learning and explain its importance in solving real-world problems.

Discuss the role of probability in machine learning. How is Bayes' theorem used in classification tasks, and why is it fundamental in probabilistic models?

Compare $L_1$ and $L_2$ regularization in machine learning. How do these techniques prevent overfitting, and what is the difference in the impact they have on model weights?

What are Autoencoders, and how are they used in unsupervised learning?

Explain the architecture of a Convolutional Neural Network (CNN).

Define one-shot learning and explain how it differs from traditional supervised learning. Discuss applications where one-shot learning is particularly beneficial.

What are some popular frameworks for implementing reinforcement learning, such as OpenAI Gym or TensorFlow Agents?

What is Q-learning, and how does it use the Q-value to guide an agent's action?

How is machine learning applied in speech processing such as speech-to-text and speaker identification?

Analyze how SVMs perform in high-dimensional spaces and discuss their applications in fields like bioinformatics.

What is $1 \times 1$ convolution in CNNs, and why is it used?

Discuss the importance of loss calculation in guiding the backpropagation and optimization process.

Discuss the role of LSTMs in natural language generation models like ChatGPT and how they handle long-term dependencies.

Discuss the rise of advanced actor-critic models with example.

Write short note on: ImageNet Competition.

Write short note on: Self Supervised Learning.

Write short note on: Generative AI.