Advanced Data Science & GenAI Master Program

Overview of AI, Generative AI, and Agentic AI.

Career paths and role mapping for AI professionals.

LinkedIn profile optimization for AI and tech hiring.

Using Python Colab for hands-on learning.

Learning workflow, tools, and program onboarding.

Variables: data types, dynamic assignment and reassignment, naming conventions, numeric types, strings, collections, and mutable vs immutable behavior.

Loops: for and while loops, range(), enumerate(), zip(), sequence traversal, and optimized iteration for data processing.

Control statements: if, elif, else, comparison/logical operators, break, continue, and real-world decision-tree logic.

Functions: reusable logic with def and return, positional and keyword arguments, and anonymous lambda functions.

Advanced Python: file handling for text, CSV, and JSON; file paths with os/pathlib; regex search, findall, sub, tokens, and groups.

Exception handling and OOP: try, except, finally, custom exceptions, classes, objects, __init__, variables, methods, inheritance, and overriding.

NumPy: arrays, arange, linspace, indexing, slicing, reshaping, broadcasting, vectorized computation, and mathematical operations.

Pandas: Series, DataFrame, CSV/Excel/JSON reads, cleaning, missing values, duplicates, casting, filtering, sorting, groupby, and feature preparation.

Visualization: Matplotlib line, bar, scatter, histogram plots; Seaborn boxplot, violinplot, countplot, pairplot, heatmap, correlation, and feature-importance visuals.

ML basics: supervised vs unsupervised learning, features, labels, train/validation/test sets, bias-variance, overfitting, underfitting, and end-to-end industry workflow.

Regression: linear regression, multiple regression with feature interactions, RMSE, MAE, and R2 score.

Classification: logistic regression, class imbalance sampling techniques, precision, recall, F1-score, and ROC-AUC.

Deep learning fundamentals: perceptron, backpropagation, optimization, and attention mechanism.

NLP fundamentals: preprocessing, Bag-of-Words, TF-IDF, Word2Vec CBOW/Skip-Gram, NER, LDA, LSA/LST, POS tagging, and evaluation metrics.

Transformer and BERT: encoder-decoder blocks, multi-head self-attention, positional encoding, MLM, NSP, and real NLP system use cases.

GenAI foundations: predictive AI vs generative AI, generative modeling, probability distributions, pre-training vs fine-tuning, and foundation models.

Use cases: chatbots, assistants, summarization, report automation, code generation, marketing content, enterprise knowledge assistants, search augmentation, and Q&A systems.

Prompt engineering: zero-shot, one-shot, few-shot, instruction design, chain-of-thought, reusable templates, evaluation, and iterative optimization.

LLMs: GPT-3.5, GPT-4, GPT-4o, LLaMA, Claude, open-source LLMs, tokenization, context windows, temperature, top-p, max tokens, hallucination, cost, and latency.

Hugging Face: Transformers library, model hub, model cards, tokenizers, pipeline APIs, and fine-tuning with Trainer.

RAG: vector databases, multimodal RAG, Graph RAG, RAG evaluation, document embedding, retrieval, and context formatting for LLMs.

LangChain and LlamaIndex: prompt templates, output parsers, chains, sequential workflows, agents, tool calling, vector DB/API integration, and retrieved-context pipelines.

Fine-tuning: full fine-tuning, PEFT, LoRA, QLoRA, instruction fine-tuning, dataset preparation, remote labeling, and overfitting mitigation.

Generative models and images: GAN, VAE, Stable Diffusion, diffusion fundamentals, DreamBooth, LoRA, Flux workflows, image conditioning, and creative generation.

LLMOps and Azure OpenAI: prompt/model versioning, monitoring, cost tracking, evaluation pipelines, deployment, Azure OpenAI architecture, endpoints, security, compliance, and app integration.

Agentic AI foundations: LLM apps vs agent systems, autonomous decision-making, goal-oriented agents, and single-agent vs multi-agent architectures.

Agent components: roles, goals, task decomposition, planning, reasoning, execution loops, tool calling, memory, feedback loops, self-reflection, failure handling, and recovery.

AutoGen: multi-agent conversation design, role-based collaboration, task delegation, message passing, and autonomous workflow use cases.

CrewAI and LangGraph: role assignment, team orchestration, state-based workflows, directed graphs, conditional routing, branching, planning, execution, retries, and error handling.

MCP, ACP, and A2A: context protocol fundamentals, agent communication, agent-to-agent interaction, distributed context sharing, secure exchange, and enterprise interoperability.

No-code/low-code agents: drag-and-drop builders, task automation, approvals, decision flows, cost and ROI evaluation, and collaboration with technical teams.

Ethics and safety: autonomous-agent risk, bias, hallucination, transparency, human-in-the-loop controls, compliance, responsible AI, and guardrails.

AWS Bedrock: foundation models, Bedrock APIs, agents, tool invocation, workflow automation, IAM, enterprise controls, and AWS service integration.

Statistical thinking: variable types, descriptive vs inferential statistics, distributions, variability, probability basics, experiments, outcomes, and events.

Probability: PDF, CDF, conditional probability, Bayes' theorem, mutually exclusive, joint, dependent, and independent events.

Sampling: population vs sample, sample size, confidence, random/systematic/cluster/stratified sampling, convenience, quota, snowball, and judgment sampling.

Statistics: central tendency, dispersion, skewness, kurtosis, ANOVA, ANCOVA, CLT, estimates, Z/T distributions, and Type I/II errors.

Hypothesis testing and EDA: null/alternative hypotheses, Z-test, t-test, chi-square, empirical rule, Z-table, normal/binomial distributions, five-point summary, box plots, outliers, IQR, and Z-score treatment.

Regression and classification: assumptions, gradient descent, train-test split, residuals, regularization, logistic regression, decision trees, random forests, KNN, Naive Bayes, metrics, cross-validation, and business classification projects.

Clustering introduction: K-means, centroid optimization, WCSS, elbow method, silhouette analysis, optimal cluster selection, and distance metrics.

Advanced clustering: hierarchical strategies, linkage criteria, DBSCAN, noise handling, K-means comparisons, and validation measures.

Ensemble learning: bagging, bootstrapping, Random Forest, feature importance, boosting, AdaBoost, gradient boosting, XGBoost, CatBoost, LightGBM, and bias-variance implications.

SVM and optimization: margin principles, support vectors, kernels, multi-class strategies, GridSearchCV, RandomSearchCV, Bayesian tuning, nested CV, curves, threshold tuning, SMOTE, SHAP, and interpretation.

Time series: trend, seasonality, stationarity, ADF, KPSS, ACF/PACF, decomposition, moving averages, exponential smoothing, ARIMA, SARIMA, Prophet, lag features, tree regressors, and walk-forward validation.

NLP: preprocessing, embeddings, Word2Vec, GloVe, FastText, Transformers, BERT, sequence-to-sequence models, classification, topic modeling, BLEU, perplexity, and other evaluation metrics.

Computer vision: image fundamentals, transformations, filtering, CNNs, classification, transfer learning, object detection, YOLO, SSD, RetinaNet, segmentation, U-Net, DeepLab, Mask R-CNN, ViT, DETR, and Swin Transformer.

Reinforcement learning: MDPs, rewards, actions, policies, Bellman equations, exploration-exploitation, Monte Carlo, TD learning, SARSA, Q-learning, policy gradients, actor-critic, DQN, PPO, DDPG, TD3, SAC, multi-agent RL, and real applications.

Model serving and packaging: FastAPI, Flask, BentoML, MLflow Models, Dockerization, and local testing for prediction behavior and latency.

Cloud deployment: AWS, Azure, S3, Blob storage, compute, IAM, SageMaker, Azure ML, CI/CD, uptime, latency, and failure monitoring.

Reproducible workflows: DVC, MLflow Registry, experiment tracking, Weights & Biases, Airflow, ZenML, Kedro, training pipelines, and inference pipelines.

Continuous training and monitoring: CT pipelines, automated deployment, GitHub Actions, Jenkins, GitLab CI, drift detection, Evidently, WhyLabs, alerts, rollbacks, and model promotion.

Excel analytics: formulas, lookup, conditional/text functions, data cleaning, validation, formatting, Power Query, descriptive statistics, filters, sorting, what-if analysis, and Goal Seek.

Excel visualization: PivotTables, PivotCharts, slicers, charts, dashboards, Power Pivot, relationships, DAX basics, add-ins, forecasting, reporting, and real analytics workflows.

Power BI: Power Query, data modeling, relationships, visual interactions, SQL/Excel connections, advanced visuals, DAX, time intelligence, dashboard design, storytelling, sharing, and customer segmentation projects.

Tableau: connections, preparation, joins, blending, pivots, SQL, visual building, maps, filters, groups, sets, parameters, drill-down, formatting, advanced charts, LOD expressions, trends, forecasting, stories, publishing, and collaboration.

SQL fundamentals: RDBMS basics, syntax, data types, tables, CRUD, querying, filtering, sorting, DISTINCT, logical operators, and result formatting.

Joins and analytics: INNER, LEFT, RIGHT, FULL, CROSS joins, UNION, subqueries, nested queries, GROUP BY, HAVING, summary stats, ranking functions, and Top-N analysis.

Advanced SQL: INSERT, UPDATE, DELETE, views, inline views, indexes, execution plans, performance tuning, normalization, procedures, functions, triggers, user-defined objects, SQL with Python, and analytical queries.

MongoDB and NoSQL: CRUD, querying, indexing, schema design, aggregation, transactions, replication, scaling, CAP, sharding, data models, analytics, caching, and distributed systems.

Hadoop fundamentals: big data concepts, distributed architecture, Hadoop master-slave structure, HDFS operations, directories, commands, permissions, quotas, MapReduce, YARN, Hive, Pig, HBase, and Sqoop.

Hadoop ecosystem: HDFS fault tolerance, YARN, MapReduce execution, SQL/NoSQL connectivity, cluster setup, configuration, scaling, and performance tuning.

Spark fundamentals: Spark features, architecture, RDDs, DAGs, transformations, actions, SparkSession, standalone mode, cluster management, RDD/DataFrame operations, interoperability, Spark SQL, and DataFrame APIs.

Spark ecosystem: Spark Streaming, DStreams, Structured Streaming, Kafka integration, MLlib, model building, evaluation, caching, partitioning, Tungsten, Catalyst, PySpark, and external data sources.

Core concepts: time and space complexity, Big-O, arrays, strings, sliding window, prefix/suffix, searching, linked lists, fast-slow pointers, cycle detection, stacks, queues, deques, and hashing.

Trees, graphs, and algorithms: BSTs, traversals, heaps, priority queues, BFS, DFS, Dijkstra, connectivity, cycle detection, topological sort, bipartite checks, MST, sorting, binary search, and search-on-answer.

Problem solving: recursion, backtracking, permutations, combinations, dynamic programming, greedy algorithms, interval scheduling, caching logic, rate limiting, pathfinding, optimization, debugging, and coding interview performance tuning.

Agile foundations: Scrum roles, ceremonies, artifacts, sprint planning, daily standups, reviews, retrospectives, Kanban, WIP limits, workflow optimization, user stories, story points, estimation, backlog prioritization, and metrics.

Jira and execution: projects, backlogs, sprints, issue tracking, Scrum/Kanban boards, dashboards, filters, reports, comments, mentions, notifications, permissions, limitations, and best practices.

Open-source and collaboration tools: OpenProject, Taiga, iceScrum, Redmine, Slack, Teams, Confluence, and communication for cross-functional AI and data science projects.