Deep Learning Specialization: Improving Deep Neural Networks: study plan and how long it takes
The DeepLearningAI playlist is 34 videos and 4.7 hours long. Here's how long it takes at different paces, a week-by-week schedule built from the real videos, and a planner that turns it into your own daily plan.
34 videos · 4.7 hours · DeepLearningAI
Course 2 of the Deep Learning Specialization covers making networks train well: train/dev/test splits, bias and variance, regularization and dropout, vanishing and exploding gradients, gradient checking, mini-batch gradient descent, momentum, RMSprop and Adam, hyperparameter tuning, batch normalization, softmax and an introduction to TensorFlow.
It's under five hours, but every video is a practical technique you'll use constantly.
Who it's for: Learners who finished Course 1 or know neural network basics and want to train models effectively.
How long does it take?
The full playlist is 4.7 hours across 34 videos. Here's roughly how many days that is at different daily watch times and speeds, counting video time only:
| Daily time | At 1x | At 1.5x | At 2x |
|---|---|---|---|
| 30 min a day | 10 days | 7 days | 5 days |
| 1 h a day | 5 days | 4 days | 3 days |
| 2 h a day | 3 days | 2 days | 2 days |
Make your own schedule
The playlist link is already filled in. Press Plan, set how much time you have each day and your speed, and you'll get a day-by-day schedule with a finish date and a calendar file. It reads the live playlist, so new videos are included.
Week-by-week schedule
A sample plan at 30 minutes a day and normal speed, using whole videos so none is split across days. That's 2 weeks in total. For a different pace, start date or speed, use the planner above.
Week 1Videos 1–24 · 3h 6m
Day 1 · 27m
- 1Train/Dev/Test Sets (C2W1L01)12m
- 2Bias/Variance (C2W1L02)9m
- 3Basic Recipe for Machine Learning (C2W1L03)6m
Day 2 · 26m
- 4Regularization (C2W1L04)10m
- 5Why Regularization Reduces Overfitting (C2W1L05)7m
- 6Dropout Regularization (C2W1L06)9m
Day 3 · 27m
- 7Understanding Dropout (C2W1L07)7m
- 8Other Regularization Methods (C2W1L08)8m
- 9Normalizing Inputs (C2W1L09)6m
- 10Vanishing/Exploding Gradients (C2W1L10)6m
Day 4 · 25m
- 11Weight Initialization in a Deep Network (C2W1L11)6m
- 12Numerical Approximations of Gradients (C2W1L12)7m
- 13Gradient Checking (C2W1L13)7m
- 14Gradient Checking Implementation Notes (C2W1L14)5m
Day 5 · 29m
- 15Mini Batch Gradient Descent (C2W2L01)11m
- 16Understanding Mini-Batch Gradient Dexcent (C2W2L02)11m
- 17Exponentially Weighted Averages (C2W2L03)6m
Day 6 · 23m
- 18Understanding Exponentially Weighted Averages (C2W2L04)10m
- 19Bias Correction of Exponentially Weighted Averages (C2W2L05)4m
- 20Gradient Descent With Momentum (C2W2L06)9m
Day 7 · 29m
- 21RMSProp (C2W2L07)8m
- 22Adam Optimization Algorithm (C2W2L08)7m
- 23Learning Rate Decay (C2W2L09)7m
- 24Tuning Process (C2W3L01)7m
Week 2Videos 25–34 · 1h 38m
Day 8 · 25m
- 25Using an Appropriate Scale (C2W3L02)9m
- 26Hyperparameter Tuning in Practice (C2W3L03)7m
- 27Normalizing Activations in a Network (C2W3L04)9m
Day 9 · 25m
- 28Fitting Batch Norm Into Neural Networks (C2W3L05)13m
- 29Why Does Batch Norm Work? (C2W3L06)12m
Day 10 · 28m
- 30Batch Norm At Test Time (C2W3L07)6m
- 31Softmax Regression (C2W3L08)12m
- 32Training Softmax Classifier (C2W3L09)10m
Day 11 · 22m
- 33The Problem of Local Optima (C2W3L10)5m
- 34TensorFlow (C2W3L11)16m
Tips for finishing
- Implement Adam and dropout from scratch once; afterwards, library versions make much more sense.
- Take notes on the hyperparameter tuning advice; it's the most reusable part of the course.