Six Month - AI/ML Preparation

Mayur Jain

profile
Six Month - AI/ML Preparation
profile
₹30,000₹36,000
1500 mins

Topic covered in 1500 Minutes


Technical Skills


  • Programming Language—Python
  • Mathematics—Linear Algebra, Calculus, Statistics.
  • Machine Learning—ML Algorithms, ML concepts,
  • Deep Learning—Basics of Neural Networks
  • Generative AI—LLMs (RAG, Agents, Prompt Engineering)
  • MLOps—Deployment, tools, and frameworks.


Programming Language - Python

  • It is simple, easy to use, and quick to learn.
  • Provides 1000+ Python packages, tools, and frameworks to handle the A to Z of machine learning, deep learning, generative AI, and MLOps.
  • Start with simple concepts like variables, conditional statements, loops, etc. Move on to the OOP (object-oriented programming) concept and build simple applications.


Mathematics

A basic understanding of mathematics is crucial for success in the ML and data science roles.


  • Linear Algebra — Vector, Matrix, Rank of a Matrix, Eigenvector, Eigenvalues.
  • Calculus — Gradient, Partial Differentiation, Chain Rule.
  • Probability and Statistics — Various probability functions, conditional probability, central limit theorem, hypothesis testing, and so on.
  • Understanding various loss functions, activation functions, and so on.


Machine Learning

The number of topics in machine learning is difficult to write down. We’ll try to make it simple by segregating ML into two parts: general ML concepts and ML algorithms.


ML Concepts include

  • Features and target
  • Train-test split of a dataset
  • Validation dataset
  • Loss function
  • Gradient Descent.
  • Regularization
  • Training and validation error
  • K-fold CV, and many more concepts.


ML algorithms include all the basic algorithms for classification, regression, and unsupervised learning.

  • Linear regression
  • Logistic regression
  • Tree-based models
  • SVM
  • Naive Bayes
  • K-means
  • K-Nearest Neighbour, and more.


Deep Learning

Deep learning is the future of AI. There are tonnes of topics to cover in DL. But for the sake of simplicity, we can start with


  • Neural networks — Node, layers (hidden, input, output)
  • Training setup.
  • Early stopping
  • How to perform inference?
  • Activation functions
  • Loss function
  • Batch Norm
  • Vanishing Gradient
  • Dropout, weight decay, and so on.
  • Batch size, epoch.
  • Gradient checkpointing.
  • DL architectures like CNN, RNN, Transformers, etc.


There are tonnes of research papers on deep learning architectures; it will be beneficial if some stand-out papers (ResNet, AlexNet, Transformers, EfficientNet, MobileNet, etc.) are deep-dived from scratch for better knowledge.


Generative AI

The hottest topic of the last two years is generative AI. To put it simply, LLMs. In-depth understanding of transformer architecture and its variants is really important.


Key topics under GenAI

  • Building RAG system
  • Building a chatbot with memory
  • Building an agentic system with multimodality.
  • Efficient Prompt and Context Engineering.
  • Retrieval and ranking aspects of RAG and other similar systems.


Implementing all the systems above and adding additional features to the system as we move towards more complex systems, like multimodal chatbots or systems.

Terms | Privacy