(Ebook) The Crystal Ball Instruction Manual Volume Two: Foundations for Data Science by Stephen Davies, Ph.D

(Ebook) The Crystal Ball Instruction Manual Volume Two: Foundations for Data Science by Stephen Davies, Ph.D

Since you’re still reading, this means you’re still interested, and
ready to explore the next level. In Foundations, we’ll solidify your
growing knowledge so that you have a firm base on which to build
everything else. Future volumes in this series will cover applications, advanced techniques, and some special data types that require special treatment. But it all ultimately rests on the skill set
you’ll have built in Volumes One and Two of this series.

We’re going to dive right in. But first, let me make a list of some
of the things I’m counting on you being pretty solid on:

• Atomic and aggregate variables – ints, floats, strings
– NumPy arrays, Pandas Serieses, DataFrames
– What all these look like in memory

• Scales of measure (especially categorical vs. numeric) 

• Assocation vs. causality
– Causal diagrams
– Dependent and independent variables
– Observational studies vs. controlled experiments
– Confounding factors

• Statistical significance 

• Exploratory Data Analysis
– Bar charts
– Histograms
– Contingency tables
– Scatterplots
– Boxplots
– Quantiles
– Outliers

• Python stuff
– Creating and performing calculations on atomic data
– Creating and accessing NumPy arrays
– for loops
– if statements (and if/elif/else)
– Functions and methods
∗ Calling a function/method vs. writing a function
∗ Passing arguments (both atomic and aggregate)
∗ Return values
∗ Function that modify in-place vs. returning a copy
– Pandas Series & DataFrames
∗ Reading from a .csv file
∗ The “index”
∗ .value_counts()
∗ .groupby()
∗ Using .iloc[] vs. .loc[] vs. []
∗ Single ints/labels vs. slices vs. lists
∗ Queries
∗ Recoding / transforming columns 

• ML concepts
– Classification vs. regression
– Features and target attribute
– Training data, test data, and new data
– Random sampling
– The “prior” vs. “posterior”
– Evaluating a classifier

All these topics were covered in Crystal Ball Volume One. If any
of them are iffy, you might take an hour or two to flip back through
them and brush up!