Descriptive Statistics

Central Tendency

• Mean: sum/count, sensitive to outliers
• Median: middle value, robust
• Mode: most frequent

Spread

• Range: max - min
• Variance: avg squared distance from mean
• Std: √variance, same units as data

Percentiles

• 25th = Q1, 50th = median, 75th = Q3
• IQR = Q3 - Q1
• Used for outlier detection

Normal Distribution

• Bell curve: N(μ, σ)
• 68% within 1σ, 95% within 2σ
• Central Limit Theorem

Correlation

• Pearson r: [-1, +1]
• Only measures linear relationship
• cov(X,Y) / (std_X * std_Y)

Z-scores

• z = (x - mean) / std
• Standardizes to mean=0, std=1
• Compare across different scales

Outliers

• |z| > 3 or IQR method
• Investigate before removing
• Could be errors or real signal