Applied Data Science
summary
Walks through installing Anaconda/Miniconda and running IPython notebooks, trains a Random Forest (100 estimators → ~90% on MNIST), then builds a Keras neural net (tuned to ~98% accuracy, GPU recommended) and shows training/prediction times. Demonstrates Word2Vec and a Word2Vec Twitter meme bot that actually generates image-text memes from a timeline. Covers practical data munging, validation metrics (R², ROC/AUC), visualization advice, and adversarial examples that break classifiers — notebooks and setup URLs provided so the code runs.
AI-written summary · about
chapters (9)
- 0:00Start
- 19:15The Cutting Edge
- 25:12Limitations and Bumps
- 26:47Tech Grab Bag
- 29:10The Magician Tutorial
- 29:53Breaking Down The Theory
- 35:47Random Forests
- 37:05Basic Neural Net
- 38:53Activation Functions
related talks
01h 38mEffective Data Visualization
02h 31mTech Tutorial: Data visualization in Python
01h 29mPandas and Plots for Data Analysis
43mVisualising Machine Learning Models
01h 00mData, visualization, and designing AI
33mData Science & Data Visualization in Python
37mUsing Exploratory Data Analysis to Discover Patterns
43mEveryone can do Data Science in Python