Data Science and Artificial Intelligence is an Advanced Level course for students who have completed the Data Analytics with Python course. After completion of the Advanced Level course, students can apply for positions as Machine Learning Engineer, Deep Learning Engineer,Computer Vision Engineer,Natural Language Processing (NLP) Engineer, Data Scientist, AI Research Scientist, AI Consultant, Academic Researcher or Educator, Startup Founder or Entrepreneur(AI focused), Quantitative Analyst (Quant) and similar.

Entry Requirement:

• Data Analysis with Python
• The selection of the candidate is subject to a comprehensive test to analyze their knowledge level.

1. Machine Learning

1.1 Understanding the fundamentals of machine learning

• Different types of machine learning: supervised, unsupervised, and reinforcement learning
• Real-world applications and examples of machine learning
• Introduction to popular machine learning libraries: scikit-learn 

1.2 Data Preprocessing and Feature Engineering I

• Handling missing data and outliers
• Data normalization and standardization
• Encoding categorical variables – ordinal encodings, one hot encodings, label encodings
• Feature scaling and selection

 1.3 Supervised Learning: Regression

• Introduction to regression problems
• Linear regression: theory and implementation in scikit-learn
• Polynomial regression and regularization techniques
• Evaluating regression models: metrics and cross-validation

1.4 Supervised Learning: Classification

• Introduction to classification problems
• Logistic regression and binary classification
• Decision trees and random forests
• Naive Bayes
• Support Vector Machines (SVM)
• K-Nearest Neighbors (KNN)
• Tree based models – XGBoost, CatBoost, LightGBM
• Ensembling models – Bagging and Boosting;Model Stacking and Blending
• Model evaluation for classification

1.5 Data Preprocessing and Feature Engineering II

• Outlier removal with advanced techniques
• Missing values imputation – advanced techniques.
• Column transformers and sk-learn pipelines.
• Feature construction and splitting

1.6 Machine Learning Concepts

• Hyper parameter tuning – Grid Search CV, Random Search CV
• Hyper parameter tuning using optuna
• Cross Validation techniques – k fold, stratified k fold
• ML Experimentations using weights and Biases

1.7 Unsupervised Learning: Clustering

• Introduction to clustering and unsupervised learning
• K-means clustering: theory and implementation
• Agglomerative Clustering
• Hierarchical clustering and density-based clustering
• Evaluating clustering results and choosing the number of clusters

1.8 Unsupervised Learning: Dimensionality Reduction

• PCA (Principal Component Analysis)
• Reducing the complexity of data and visualizing high-dimensional data
• Applications of dimensionality reduction in feature selection and compression
• Linear Discriminant Analysis (LDA)
• t-Distributed Stochastic Neighbor Embedding (t-SNE)

2. Deep Learning

2.1 Artificial Neural Networks

2.1.1 Introduction to Deep Learning

• Understanding the basics of neural networks and deep learning
• Historical context and key milestones in deep learning
• Introduction to popular deep learning frameworks: TensorFlow,Keras,pyTorch
• Setting up the development environment for deep learning

2.1.2 Mathematical Concepts for Deep Learning

• Probability and Statistics
• Linear Algebra
• Calculus – Applied Derivatives and Integration
• Information Theory

2.1.3 Neural Networks and Backpropagation

• Perceptron – introduction, concepts, limitations
• Introduction to Multi Layer perceptron.
• Review of basic neural network architecture
• Activation functions and their role in neural networks
• Advantages and limitations of Activation Functions
• Gradient Descent Optimizations and its types
• Forward propagation and backpropagation algorithms
• MLP Memoization 
• Vanishing and exploding gradients problem

2.1.4 Neural Networks Concepts

• Early Stoppings
• Drop out layers
• Batch Normalization
• Regularization techniques (L1 and L2)
• Optimizers in Neural networks
• Hyper parameters in neural networks
• Hyper parameter tuning in NN

2.1.5 Hands on project  – Classification/Regression using ANN

3. Computer Vision 

 3.1 Introduction to Convolutional Neural Networks (CNNs)

• Basics of image data 
• Understanding convolutional layers and their role in feature extraction
• Filters, paddings and strides
• Pooling layers for  down-sampling and spatial hierarchy
• Image classification using CNN
• Project – Building a simple image classification model using CNNs

3.2 CNN architectures

• Introduction to popular CNN architectures (LeNet, AlexNet, VGG, ResNet,EfficientNet etc.)
• Understanding the design principles behind each architecture
• Transfer learning and fine-tuning pre-trained CNN models
• Intro to the application attention mechanisms in CNN(e.g., self-attention)
• Case studies of CNNs applied to various domains

3.3 Image Preprocessing and Augmentation

• Data preprocessing techniques (rescaling, normalization, etc.)
• Data augmentation for expanding training datasets
• Handling imbalanced datasets and class distribution issues
• Implementing image preprocessing pipelines using libraries like TensorFlow/Keras

3.4 Object Detection and Localization

• Introduction to object detection and localization tasks
• R-CNN, Fast R-CNN, and Faster R-CNN architectures
• Single Shot MultiBox Detector (SSD) and You Only Look Once (YOLO)
• Implementing object detection models and applications
• Hands on project related to Object detection

3.5 Image Segmentation and Semantic Segmentation

• Understanding image segmentation and its applications
• Fully Convolutional Networks (FCN) for semantic segmentation
• U-Net architecture and instance segmentation
• Applying segmentation models for medical and satellite image analysis
• Hands on project related to Image Segmentation

3.6 Project Development and deployment using Streamlit/ Gradio

4. Natural Language Processing (NLP)

 4.1 Introduction to Natural Language Processing

• Understanding the scope and applications of NLP
• Key challenges and nuances in processing human language
• Introduction to popular NLP libraries and tools (NLTK, spaCy, etc.)
• Setting up the development environment for NLP tasks

4.2 Text Preprocessing and Tokenization

• Cleaning and normalizing text data
• Tokenization and sentence segmentation
• Stop-word removal and stemming/lemmatization
• Handling special characters and encoding issues

4.3 Part-of-Speech Tagging and Named Entity Recognition

• Understanding grammatical roles with part-of-speech(POS) tagging
• Identifying and extracting named entities (NER)
• Leveraging pre-trained models for NER
• Applications of NER in information extraction

4.4 Sentiment Analysis and Text Classification

• Introduction to sentiment analysis and its applications
• Building sentiment analysis models using machine learning
• Leveraging deep learning for sentiment analysis
• Case studies and real-world examples of sentiment analysis
• Hands on project for sentiment classification.

 4.5 Language Modeling and Text Generation

• Basics of language modeling and n-grams
• Introduction to Markov models and hidden Markov models
• Generating text using Markov chains and n-grams
• Introduction to recurrent neural networks (RNNs) for text generation

4.6 Word Embeddings and Word2Vec

• Understanding word embeddings and distributed representations
• Word2Vec algorithm and continuous bag-of-words (CBOW) model
• Skip-gram model for learning word vectors
• Using pre-trained word embeddings for various NLP tasks

4.7 Sequence Models

• Basics of sequential data and its challenges
• Understanding recurrent neural networks (RNNs)
• Applications of RNNs in sequence modeling (e.g., language generation, translation)
• Introduction to Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU)
• Limitations of RNNs, LSTMs and GRU’s for sequence models
• Self attention mechanisms and Transformer architecture
• Project related to sequence models using all the algorithms mentioned above.

4.8 Named Entity Recognition and Relation Extraction

• Deep dive into Named Entity Recognition (NER) techniques
• Relation extraction and entity linking
• Building end-to-end NER and relation extraction pipelines
• Applying NER and relation extraction in information retrieval(project)

4.9 Advanced NLP Topics and Future Trends

• Transfer learning and pre-trained language models (e.g.,DeBERTa, RoBERTa,  BERT, GPT-3)
• Handling large text corpora and big data in NLP
• Ethical considerations in NLP and bias detection
• Recent advancements and ongoing research in NLP 
• Hands on project related to fine tuning of models.

5. Generative Models and LLMs

• Introduction to Generative AI and LLMs
• Understanding generative vs. discriminative models
• Generative Adversarial Networks (GANs) with hands-on projects.
• Overview of LLMs and their significance
• Pre-training and fine-tuning of LLMs (project)
• Recent developments and research in Generative AI
• Recent developments and research in LLMs.

Course Learning Outcomes

The students will be able to: 

• Gain a holistic grasp of machine learning, deep learning, computer vision, and NLP for adept data analysis and AI utilization.
• Master foundational concepts, including supervised and unsupervised learning, regression, and classification techniques.
• Develop proficiency in implementing advanced methods like CNN architectures, image preprocessing, and object detection.
• Acquire skills in NLP essentials such as text preprocessing, sentiment analysis, and sequence models.
• Explore cutting-edge domains, like transfer learning, GANs, and ethical considerations in AI and NLP.
• Stay abreast of industry trends and ethical implications while delving into generative models and emerging research.
• Emerge with a robust toolkit to solve complex data challenges and contribute to the forefront of AI innovation.

Career Opportunities

• Machine Learning Engineer
• Deep Learning Engineer
• Computer Vision Engineer
• Natural Language Processing (NLP) Engineer
• Data Scientist
• AI Research Scientist
• AI Consultant
• Academic Researcher or Educator
• Startup Founder or Entrepreneur(AI focused)
• Quantitative Analyst (Quant) and so on.