---

date: 2026-02-07

📊 Statistics & Probability Roadmap for Data Science Freshers

This roadmap covers everything a Data Science / Machine Learning fresher must know in Statistics and Probability — from absolute basics to interview-critical concepts.

The focus is on understanding + application, not formula memorization.

---

🟢 LEVEL 0: Absolute Foundations (Must-Know)

Descriptive Statistics

• Mean
• Median
• Mode
• Range

Measures of Spread

• Variance
• Standard Deviation
• Interquartile Range (IQR)

Key understanding: - Effect of outliers on mean vs median - When to prefer median over mean

---

🟢 LEVEL 1: Data Shape & Distribution

Distribution Basics

• Data distribution
• Frequency distribution
• Probability distribution (conceptual understanding)

Normal Distribution

• Bell curve
• Mean, median, mode in normal distribution
• 68–95–99.7 rule

Density Curve

• Interpretation of density plots
• Relationship between area and probability

---

🟢 LEVEL 2: Skewness & Outliers

Skewness

• Symmetric distribution
• Right-skewed (positive skew)
• Left-skewed (negative skew)

Outliers

• What are outliers
• IQR method for outlier detection
• Z-score method for outliers
• Impact of outliers on:
  • Mean
  • Variance
  • Machine Learning models

---

🟡 LEVEL 3: Probability Fundamentals

Basic Probability

• Random experiment
• Sample space
• Event
• Probability rules
  • Addition rule
  • Multiplication rule

Conditional Probability

• Definition of conditional probability
• Formula: P(A|B)
• Real-world intuition (medical testing, spam detection)

---

🟡 LEVEL 4: Bayes’ Theorem

• Bayes’ theorem formula
• Intuition behind Bayes’ theorem
• Prior, likelihood, posterior
• Real-world examples
• Importance in Naive Bayes algorithm

---

🟡 LEVEL 5: Random Variables & Probability Distributions

Random Variables

• Discrete random variable
• Continuous random variable

Probability Distributions

• Bernoulli distribution
• Binomial distribution
• Poisson distribution
• Normal (Gaussian) distribution

Understanding focus: - When to use each distribution - Real-world examples for each

---

🟠 LEVEL 6: Z-Score & Data Scaling

Z-Score

• Z-score formula
• Interpretation of Z-score
• Z-score for outlier detection

Data Scaling

• Standardization
• Normalization
• Difference between standardization and normalization
• Why scaling is required in ML
• Models sensitive to scaling:
  • KNN
  • SVM
  • Linear Regression
• Models not sensitive to scaling:
  • Decision Trees
  • Random Forest

---

🔴 LEVEL 7: Inferential Statistics

Hypothesis Testing

• Null Hypothesis (H₀)
• Alternative Hypothesis (H₁)
• Significance level (α)
• p-value
• Decision rule in hypothesis testing

Errors in Hypothesis Testing

• Type I Error (False Positive)
• Type II Error (False Negative)
• Real-world examples of both errors

---

🔴 LEVEL 8: Statistical Tests

Types of Tests

• One-tailed test
• Two-tailed test

Common Statistical Tests

• Z-test
• T-test
  • One-sample t-test
  • Two-sample t-test
• ANOVA (Analysis of Variance)
  • One-way ANOVA
  • Why ANOVA is preferred over multiple t-tests

---

🔴 LEVEL 9: Correlation & Relationship Analysis

• Covariance
• Correlation
• Pearson correlation coefficient
• Spearman rank correlation
• Correlation vs causation
• Multicollinearity (basic understanding)

---

🟣 LEVEL 10: Statistics for Machine Learning Thinking

Advanced Concepts (Fresher Advantage)

• Bias vs Variance tradeoff
• Sampling vs Population
• Central Limit Theorem (intuition level)
• Confidence Interval (conceptual understanding)

---

🎯 Final Notes

• Levels 0–7 are mandatory for interviews
• Levels 8–10 provide a strong competitive edge
• Focus on:
  • Intuition
  • Practical examples
  • Code-based validation
• Avoid memorizing formulas blindly

Statistics is the foundation of Machine Learning.
Strong statistics = confident ML decisions.