Intro to Machine Learning

Machine Learning

Is about extracting knowledge from data.

Machine Learning

It is predictive analytics or statistical learning.

Supervised Learning

Develop predictive model based on both input and output data.

Supervised Learning

User provides the algorithm with pairs of inputs and desired outputs.

Supervised Learning

The algorithm is able to create an output for an input it has never seen before without any help from a human.

Examples of Supervised Machine Learning

• Identifying the zip code from handwritten digits on an envelope.

• Determining whether a tumor is benign based on a medical image.

• Detecting fraudulent activity in credit card transactions.

• Spam classification.

Unsupervised Learning

Group and interpret data based only on input data.

Unsupervised Learning

Only the input data is known, and no known output data is given to the algorithm.

Unsupervised Learning

Harder to understand and evaluate.

Examples of Unsupervised Machine Learning

• large collection of text data.

• Segmenting customers into groups with similar preferences.

• Detecting abnormal access patterns to a website.

Automate decision-making processes

A successful machine learning is?

A representation of your input data that a computer can understand.

What is needed to both supervised and unsupervised machine learning tasks?

Entry point

Represented by an individual entity together with its properties.

Samples, Features

rows are the ——, columns are the ——

Essential Libraries and Tools for Machine Learning in Python

• Jupyter Notebook through Anaconda

• Google Colab

• Pycharm

• Spyder

Esssential Libraries

• Scikit learn

• Numpy

• SciPy

• matplotlib

• pandas

Scikit learn

Contains machine learning algorithms.

Numpy

Functionality for multidimensional arrays, linear algebra operations, Fourier transform, and pseudorandom number generators.

SciPy

Advanced linear algebra routines, mathematical function optimization, signal processing, special mathematical functions, and statistical distributions.

Matplotlib

Is the primary scientific plotting library in Python, publication-quality visualizations such as line charts, histograms, scatter plots, and so on.

Pandas

Is a Python library for data wrangling and analysis. It is built around a data structure called the DataFrame.

Types of Supervised Machine Learning

1. Classification tasks

Classification tasks

The goal is to predict a class label, which is a choice from a predefined list of possibilities.

Types of Classificatoin

1. Binary - distinguish between 2 classes. (Yes or no,

   positive or negative, spam or not)

2. Multiclass classification - classification between

   more than two classes.

Regression tasks

• to predict a continuous number.

• person’s annual income from their education and their age.

• Predicted value – amount of income, any numeric value

Generalization

If a model can make accurate predictions on unseen data.

Overfitting

Occurs when you fit a model too closely to the particularities of the training set and obtain a model that works well on the training set but is not able to generalize to new data.

Underfitting

Refers to a model that can neither model the training data nor generalize to new data.Preprocessing Data Using Different

Technique

Preprocessing Data Using Different Technique

1. Data Normalization

2. Data Standardization

3. Label Encoding

Data Normalization

Is a rescaling of the data from the original range so that all values are within the new range of 0 and 1.

Data Standardization

Scales each input variable separately by subtracting the mean (called centering) and dividing by the standard deviation to shift the distribution to have a mean of zero and a standard deviation of one.

Data Standardization

Involves rescaling the distribution of values so that the mean of observed values is 0 and the standard deviation is 1.

Label Encoding

Transforming the word label into numerical form so that the algorithm can understand how to operate them.

Label Encoding

Transforming the word label into numerical form so that the algorithm can understand how to operate them.

fit() function

Using training data to estimate the minimum and maximum observable values with?

transform() function

Use the normalized data to train your model. This is done by calling?