Data Science – End 2 End Beginners Course Part 1

This lecture will cover Python programming language. We will discuss why python is popular for data science, how to learn python and a basic example to show it's syntax. Also, showing you the directory structure of the sample python code in github.

This lecture will cover the data type numbers in python.

This lecture will cover the data type variables in python.

This lecture will cover the data structures - Lists and Arrays. Lists are the most common. Arrays are not native to python but can be created using lists

This lecture will cover the data structures - Sets, Tuples and Dictionary. Dictionaries are widely used in python

This lecture will cover the data type strings in python.

This lecture will cover the different types of python operators - maths, assignment, comparison and logical. They are used for performing different types of operations on different data types.

In this lecture we will learn the different control flow statements in python - if else, while, for, range and list comprehension.

This lecture will cover the basics on functions and error handling

This lecture will teach how to Read/Write from/to different sources - console, files, databases

This lecture will teach the basics on Object Oriented Programming principles, to show how to use them to build modular machine learning programs. It is not a full OOPs programming course.

In this lecture we will show how to create modules and packages

In this lecture we will cover some useful general programming concepts - sort algorithms, recursion and regex. They are widely used in programming, so its beneficial to learn about them. We will cover the basics (not all the different types of each).

This lecture will give an introduction about the data analytics library pandas and its use in data science

In this lecture we will learn about the basic pandas data structures - series and dataframes, how to create them and manipulate them using some basic operations.

This lecture will show how to view useful information and perform slicing & conditional operations on series and dataframes. It will also show how to apply functions to the entire dataframe in a single operation.

This lecture will teach you how to handle missing or junk values in the data. It will also show how to combine multiple dataframes using merge and concat.

This lecture will cover text data (strings) and how it is can be processed. How to iterate over series and dataframes. And it will also cover data sorting methods.

This lecture will teach you how to group dataframes across different labels or levels and apply transformation and aggregate functions. It will also cover type casting or conversion from one data type to another.

This lecture will teach you how to handle and manipulate time series data. It will also cover resampling and rolling window techniques to change the frequency of the data.

In this lecture we will learn about reading and writing data from/to different format types (csv, json,etc). We will also learn how to plot data using different chart types.

In this lecture we will show you how to use pandas to perform basic SQL query operations.

In this lecture series students will learn the basic concepts in statistics required for data science.

In this lecture we will learn about the basic statistics that describe a samples central tendency and variability.

This lecture will cover the basics on distributions - histograms and normal distribution

In this lecture students will learn more details about distributions - skewness, tails, outliers and kurtosis

In this lecture we will learn about more ways to analyse single variable distributions using descriptive statistics - percentile, range, quartile and boxplot.

This lecture will cover inferential statistics - point and interval estimates. They are used for deducing properties of the population from the sample.

This lecture will cover inferential statistics - hypothesis testing. It is used to test an assumption about a population parameter using a test statistic

This lecture will cover bivariate analysis techniques - correlation and covariance

This lecture will teach another bivariate analysis technique - regression. It is one of the most widely used techniques for building machine learning models.

In this lecture we will cover some more useful statistics concepts - exponentials and logarithms. They are used for different machine learning models depending on the data distribution.

This is an introduction to the probability lecture series

This lecture will cover the basic terminology and axioms in probability, necessary to understand the basics of probability required for data science.

This lecture will cover Conditional probability and Law of Total probability. They serve as the basic foundation to understand some machine learning models.

This lecture will cover Bayes Theorem. It is used for different machine learning algorithms.

In this lecture we will learn about some more useful probability concepts - Central Limit Theorem and Probability Mass Function (PMF).

This lecture will cover the basic terminology in sampling, different types of sampling, sampling errors and biases.

This lecture will give a high level introduction and process flow diagram for solving data science problems.

This lecture will cover how to describe the business problem, perform EDA (Exploratory Data Analysis) and provide possible solutions.

This lecture will show how to select the right algorithm(s) for the problem based on the prediction type, problem type and sample size.

This lecture will teach you how to perform basic data wrangling, splitting the sample into training and test set and how to standardize the data.

This lecture will teach you how to build, evaluate and select the best model using Pipeline and GridSearchCV libraries with hyperparameter tuning.

This lecture will cover how to visualize the data using different libraries and briefly cover how to deploy and retrain/redeploy the model on a local or cloud platform.

This lecture will cover how to perform data wrangling including cleaning, formatting, encoding, binning and aggregating the data.

This lecture will give an introduction to Regression algorithms, Error Metrics used to measure their accuracy and the Bias-Variance Tradeoff. In this lecture series you will learn how to build regression algorithms using python, pandas, scikit learn and other libraries. They will go through the entire data science process, step by step, creating a template that can be used for most models. You will see that building the model requires only a few lines of code. But there is a lot of background learning that is required to build, explain and improve them.

In this lecture you will learn how to build a linear regression model using python, pandas and scikit learn libraries.

In this lecture you will learn how to build a multiple linear regression model using python, pandas and scikit learn libraries. It will show that adding more useful data (X independent variables) improves the results because the model has more information to learn from. Students will now see the benefit of using the data science process and code template, allowing them to build the algorithms efficiently.

In this lecture you will learn how to build a polynomial regression model using python, pandas and scikit learn libraries. They create a polynomial function of the independent variables X (higher order - X^2, X^3, etc). It will show the disadvantages of using non-linear models for linear equations.

In this lecture you will learn how to build regularization (L1 Lasso and L2 Ridge) regression models  using python, pandas and scikit learn libraries. They add a penalty term to the loss function to reduce the model complexity.

In this lecture you will learn how to use decision trees to build regression models using python, pandas and scikit learn libraries. It will explain how the trees are built, metrics used (Entropy, Information Gain, Gini Impurity), libraries for plotting trees (Graphviz) and how they give better results compared to multiple linear regression.

In this lecture you will learn about the different types of Ensemble algorithms (Boosting and Bagging, Random Forest, XGBoost, etc) used to build machine learning models.

In this lecture you will learn how to use Random Forest algorithm to build regression models using python, pandas and scikit learn libraries. It will explain how they use multiple decision trees, metrics used, different parameters that can be optimized, how to use feature importance for feature selection and how they give better results compared to decision trees.

In this lecture you will learn how to use XGBoost to build regression models using python, pandas, scikit learn and xgboost libraries. It will explain how it uses multiple gradient boosted decision trees, metrics used, different parameters that can be optimized, how to use feature importance for feature selection and how they give better results compared to decision trees.

In this lecture you will learn how to use Pipeline and GridSearchCV libraries to build, optimize and evaluate multiple regression algorithms at once and then select the best one. Pipeline combines data transformations and training multiple models into a sequence of steps. GridSearchCV performs k-fold cross-validation to train models on multiple variations of the training data.

In this lecture series you will learn about another type of supervised learning algorithms - classification. Classification algorithms are used for predicting a class or category for observations in the data. Error metrics like confusion matrix and ROC curve are explained. A brief overview of the data samples that will be used for the classification problems will be given

In this lecture you will learn about building Logistic Regression Classification models using python, scikit learn and pandas libraries. It is a linear model and uses a logistic function. The lecture also covers the maths, statistics and probability underlying the logisitc function.

In this lecture you will learn about building K-NN (K-Nearest Neighbors) Classification models using python, scikit learn and pandas libraries. It is an instance based non-parametric non-linear model.

In this lecture you will learn about building SVM (Support Vector Machine) Classification models (SVC) using python, scikit learn and pandas libraries. It can be used for both linear & non-linear problems and uses the concept of hyperplanes and kernel functions. Explanation is also given for binary and multiclass classification models.

In this lecture you will learn about building different Naive Bayes Classification models (Gaussian, Multinomial, Bernoulli) using python, scikit learn and pandas libraries. It will discuss the naive bayes theorem (already covered in detail in the probability chapter) along with Maximum a Posteriori Probability (MAP) estimation and argmax function.

In this lecture you will learn about building Decision Tree, Random Forest and XGBoost Classification models using python, pandas, scikit learn and xgboost libraries. It will briefly discuss the different models, as their concepts and maths are already covered in detail in the regression chapter.

In this lecture you will learn about building Multiclass Classification models using the SVC algorithm by using python, pandas and scikit learn libraries. Multiclass classification is used when there are more than 2 output class labels (0,1,2,...,n).

In this lecture you will learn how to use Pipeline and GridSearchCV libraries to build, optimize and evaluate multiple classification algorithms at once and then select the best one. Pipeline combines data transformations and training multiple models into a sequence of steps. GridSearchCV performs k-fold cross-validation to train models on multiple variations of the training data.