🧠 Deep Learning (DLE) – VII Semester Survival Kit

Because why just suffer in class when you can suffer here too?

---

📚 Course Info (aka: What we signed up for)

• Semester: VII
• Branches: CSE, CST, CSBS, CSE(AIML), CSE(DS) (basically everyone gets dragged in)
• Course Code: CS407
• Category: PEC-IV
• Credits: 3 (but feels like 30)
• Scheme: 2020 (and hopefully future-proof 🤞)
• Marks Split:
  • Internal: 40 (a.k.a. “write anything in exam, you’ll still get some marks”)
  • External: 60 (a.k.a. “pray to all gods”)
  • Total: 100

---

🎯 Course Outcomes (COs) – Fancy way of saying "what they expect us to magically know"

• CO1: Understand concepts of Deep Learning and ANN.
• CO2: Pretend to summarize Deep Neural Nets.
• CO3: Understand CNN operations (basically filters & magic).
• CO4: Memorize CNN Architectures (LeNet, AlexNet, ResNet – Pokémon evolution line).
• CO5: Fight with RNNs and survive LSTMs.

---

📝 Units Breakdown (with sarcasm included)

UNIT – I

• Deep Learning intro (ML vs DL – the eternal “why so deep?” question).
• ANN: From neurons to “Perceptrons” (a.k.a. dot product with attitude).
• Fine-tuning hyperparameters (aka endless trial & error).
• Case Study: Heart Disease Prediction (because why not start with saving lives?).

UNIT – II

• Deep Neural Networks – training them without exploding/vanishing gradients (spoiler: good luck).
• Faster optimizers (SGD isn’t enough, now we bring Adam to the party).
• Regularization: AKA “Don’t let your model cheat by memorizing.”

UNIT – III

• CNN Part 1: Convolutions, pooling, structured outputs.
• Applications: Image recognition, object detection, and memes classification (probably).

UNIT – IV

• CNN Part 2: Architectures like LeNet5, AlexNet, GoogLeNet, ResNet.
• Advantages of CNN (hint: they actually work).
• Case Study: Handwritten digit recognition (MNIST – the “Hello World” of DL).

UNIT – V

• RNNs: From loops to memory cells (aka models that actually “remember”).
• LSTM: Because vanilla RNNs forget faster than students after exams.
• Case Study: Time series prediction (stocks, weather, or just your GPA).

---

📖 Textbooks That Everyone Buys But Reads PDFs Instead

1. Hands-On Machine Learning with Scikit-Learn and TensorFlow – Aurélien Géron (a.k.a. the only book you’ll actually read).
2. Deep Learning – Ian Goodfellow, Yoshua Bengio, Aaron Courville (a.k.a. The Holy Bible of DL).

---

📂 What’s in this Repo?

• ✍️ My personal notes (written with love, sweat, and caffeine).
• 💻 Code snippets for each unit (so you don’t cry while coding from scratch).
• 🎓 Case studies implemented in Python/TensorFlow/Keras.
• ⚡ Bonus: Extra examples, experiments, and probably some meme references.

---

🚀 Why this repo?

Because Deep Learning is deep… and so are the nightmares during exams.
This repo = Notes + Codes = Survival kit for CS407.

---

🌟 Contributions

• If you’re in the same boat (a.k.a same course), feel free to add notes/code.
• PRs welcome, but don’t send your entire syllabus PDF 😅.

---

🏁 TL;DR

This is my Deep Learning (DLE) – VII Semester repo with notes + code for each unit, case studies, and extra stuff.
Use it wisely. Or just stare at the repo and pretend to study.