- Empirical Rule: most of the data lies within 3 std deviation of normal distribution data ie. 99.7%.
- classical probability: like coin, card since we know all possible outcomes
- empirical probability: based on historical data or experiment
- subjective probability: works on belief
- even odds: like coin flip
- weighted odds: like weather forecast
- permutation: order matters
- combination: order not matters
- percentile: 98th percentile means we lie in 2% topper
- Percentile(x) = (Number of values fall under ‘x’/total number of values) × 100
- probability tree used for conditional probablilty
- dependent event: picking a two cards w/o replacement
- independent event
- Bayes Theorem
- Random Variable (discrete or continuous)
- Binomial
- Bell shaped curve
- z-score
- Random Sample
- Systematic Sample
- Source of bias
- Opportunity Sample
- Stratified Sample
- Cluster Sample
- Central Limit Theorem: sample mean is approx to the population mean
- Standard Error
- Hypothesis Testing
- ANOVA
- t-test
- goodness-of-fit test
- Pandas
- Numpy
- Matric multiplication using np.dot(mat1, mat2)
- Matplotlib
- Web Scraping using BeautifulSoup
- Intro to NLP using nltk
- Plotly interative visualization
- Machine Learning 101
- Linear regression
- Logistic regression
- Clustering models: K-means and hierarchal models
- Dimension reduction methods
- Association rules
- Ensembles methods
- Introduction to neural networks
- Decision tree models
- scikit-learn