stats-basic.md

Contents

---

Basic Statistics

It is good to refresh some stats for ML.

Prerequisites

In order to do these exercises, you need to learn the stats functions in your ML platform (Python, Java, R ..etc.).
For Python it is Numpy.
Refer to our Numpy section to learn the basics of Numpy.

Part 1 - Recommended

1.1 - Resources

Generic

• Online lecture notes : A nice series by Ellen Venable from Cabrillo College.
  Has very engaging graphics and very easy to read and understand. Each chapter has good set of exercises too.
  • CH 01 - Basic Intro - Introduces concepts like population vs. sample and variable types
  • CH 02 - Organizing and Graphing Data - Introduces basic graph types, frequency distribution.
  • CH 03 - Inspecting numerical data - Talks about numeric data analysis : mean, median, range, variance, standard deviation ..etc
• Book : 'Practical Statistics for Data Scientists' by Andrew Bruce, Peter Bruce (Safari link, Amazon Link)
  This is a fantastic book that provides concise, easy to understand explanation of essential statistics needed for Machine Learning.
  • Ch 1 : Exploratory Data Analysis
  • Ch 2 : Data and Sampling Distributions

Covariance / Correlation

• Covariance : https://www.youtube.com/watch?v=xGbpuFNR1ME (excellent video by Bradon Foltz)
• Correlation : https://www.youtube.com/watch?v=4EXNedimDMs (excellent video by Bradon Foltz)
• Covariance Matrix : https://www.youtube.com/watch?v=locZabK4Als (excellent video by Bradon Foltz)

1.2 - Check List

You should be familiar with the following

• Population vs. sample
• Variable types (quantitative, discrete, continuous)
• Plot basic graphs
• Do numerical analysis such as mean, median, variance, standard deviation
• Correlation

1.3 - Quick Exercise

This is a simple exercise designed to reinforce your learning so far.

We have some sample salary data (in thousands) from two cities.

city1 = [15,12, 20, 25, 50, 35, 75, 80, 60, 45, 36]
city2 = [40,42, 45, 60, 55, 52, 56, 52, 62, 57, 48]

Calculate mean, median, variation, standard deviation for both city data.

Part 2 - [Optional] Advanced

2.1 - Resources

Generic

• Book : 'Practical Statistics for Data Scientists' by Andrew Bruce, Peter Bruce (Safari link, Amazon Link)

  • Chapter 3 : Statistical Experiments and Significance Testing

Degrees of Freedom

• http://blog.minitab.com/blog/statistics-and-quality-data-analysis/what-are-degrees-of-freedom-in-statistics

Bootstrapping

• https://sebastianraschka.com/blog/2016/model-evaluation-selection-part2.html

Cross Validation / k-fold validation

• https://machinelearningmastery.com/train-final-machine-learning-model/
• https://stats.stackexchange.com/questions/52274/how-to-choose-a-predictive-model-after-k-fold-cross-validation
• https://sebastianraschka.com/blog/2016/model-evaluation-selection-part3.html

2.2 - Check List

You should know the following concepts

• Hypothesis testing
• p-value
• t-test
• Degrees of freedom
• ANOVA
• Chi-Square Test
• Bootstrapping

---

Contents