IITM BS Data Science and AI Lab (BSDA4001): Syllabus and Tips

Data Science and AI Lab (BSDA4001) is a 4 credit lab elective at the degree level of the IITM BS in Data Science. The handbook places it at level 4. You need Deep Learning (BSCS3004) before you take it. It is fully hands-on: each week brings a new tool or idea and an end-to-end assignment built around it.

The course page does not name an instructor. The course page and the handbook agree on the credits and the prerequisite.

What you work on each week

The page says the course aims to close the gap between classroom learning and what industry expects. The 12 weeks fall into five blocks.

• Weeks 1 to 3: a quick refresher. The Python data stack and Git. Then scikit-learn models, pipelines, tuning and metrics. Then building and training deep networks in PyTorch and TensorFlow, including saving models and training on a GPU.
• Weeks 4 and 5: images and text. Loading and preparing images with OpenCV and PIL, and transfer learning with ResNet and MobileNet. Then BERT and RoBERTa through Hugging Face: tokenising, fine-tuning and running the model.
• Weeks 6 to 8: building with LLMs. Prompt design, LangChain and calling models through the OpenAI, Gemini and Ollama APIs. Then retrieval-augmented generation with vector databases such as FAISS, Pinecone and ChromaDB. Then simple agents that use tools and plan several steps, with LangChain and LangGraph.
• Weeks 9 and 10: dashboards and forecasting. Dashboards in Power BI, Tableau, Streamlit or Plotly, and telling a story with data. Then time series work with Prophet and DeepAR, covering trend, seasonality and anomalies.
• Weeks 11 and 12: cloud and responsibility. An end-to-end pipeline on Google Cloud with BigQuery and Vertex AI, plus the ideas behind Spark, Hadoop and Airflow. The last week covers SHAP and LIME for explaining predictions, privacy and fairness, and a mini-project.

How the course is assessed

The handbook describes this lab as completely hands-on, with weekly graded assignments plus a project and a viva. It also adds industry topics meant to prepare you for job interviews. There is a limit on how many students can enrol, because of the project part.

Where it counts

• BS level stream rule. The BS level needs 2 data science stream courses at level 4 or higher. This lab is level 4 and tagged BD, which appears to mean the data science stream. See the BS degree level.
• Minors. It is not in any minor list in the handbook.
• Terms. The DS course table (updated 18 March 2026) marks it for May 2026, September 2026 and January 2027.

Its sister course on the programming side is the Application Development Lab (BSCS4010).

How to plan a lab term

A weekly lab punishes cramming. Each assignment is built around that week's topic, so a missed week is hard to recover.

• Block fixed hours for the lab every week, even in weeks with quizzes in other courses.
• Set up your environment before week 1: Python, Git, Jupyter and a GPU notebook. Create any cloud accounts early.
• The LLM weeks use outside APIs. Check early what the course provides and whether any service needs a key or a paid plan.
• Keep each assignment in its own folder in one Git repository, with a short README. This becomes your viva prep and your portfolio.
• Start the week 12 mini-project idea by week 9, so you are not choosing a topic in the last week.

Academic integrity

Every assignment, the project and the viva are meant to show your own skill. Using official documentation and tutorials to learn is normal. Copying a batchmate's notebook, or pasting AI-generated code you cannot explain, is not. The viva exists to check that you understand what you submitted. If you used an outside code snippet, note where it came from.

Common questions

Why is Deep Learning a prerequisite if week 3 revises it?

Week 3 is a quick refresher on training networks in code, not a first course. Deep Learning (BSCS3004) gives you the theory the lab assumes. See Deep Learning (BSCS3004).

Is there a quiz or end term exam in this lab?

The handbook describes the lab through weekly graded assignments, a project and a viva. Neither source lists quizzes or an end term exam for it, though the course page also points to the programme's standard course structure. Confirm the grading formula with the course team in week 1.

Can I show lab work in job interviews?

Yes, if it is your own. Well documented assignments make good portfolio pieces. See portfolio projects for IITM BS students.