Introduction
|
Machine learning is a set of tools and techniques to find patterns in data.
Some machine learning techniques are useful for predicting something given some input data.
Some machine learning techniques are useful for classifying input data and working out which class it belongs to.
Artificial Intelligence is a broader term that refers to making computers show human like intelligence.
Some people say Artificial Intelligence to mean machine learning
All machine learning systems have some kinds of limitations
|
Regression
|
We can model linear data using a linear or least squares regression.
A linear regression model can be used to predict future values.
We should split up our training dataset and use part of it to test the model.
For non-linear data we can use logarithms to make the data linear.
|
Introducing Scikit Learn
|
Scikit Learn is a Python library with lots of useful machine learning functions.
Scikit Learn includes a linear regression function.
It also includes a polynomial modelling function which is useful for modelling non-linear data.
|
Clustering with Scikit Learn
|
Clustering is a form of unsupervised learning
Unsupervised learning algorithms don’t need training
Kmeans is a popular clustering algorithm.
Kmeans struggles where one cluster exists within another, such as concentric circles.
Spectral clustering is another technique which can overcome some of the limitations of Kmeans.
Spectral clustering is much slower than Kmeans.
As well as providing machine learning algorithms scikit learn also has functions to make example data
|
Neural Networks
|
Perceptrons are artificial neurons which build neural networks.
A perceptron takes multiple inputs, multiplies each by a weight value and sums the weighted inputs. It then applies an activation function to the sum.
A single perceptron can solve simple functions which are linearly separable.
Multiple perceptrons can be combined to form a neural network which can solve functions that aren’t linearly separable.
We can train a whole neural network with the back propagation algorithm. Scikit-learn includes an implementation of this algorithm.
Training a neural network requires some training data to show the network examples of what to learn.
To validate our training we split the the training data into a training set and a test set.
To ensure the whole dataset can be used in training and testing we can train multiple times with different subsets of the data acting as training/testing data. This is called cross validation.
Deep learning neural networks are a very powerful modern technique. Scikit learn doesn’t support these but other libraries like Tensorflow do.
Several companies now offer cloud APIs where we can train neural networks on powerful computers.
|
Ethics and Implications of Machine Learning
|
Machine learning is often thought of as unbiased and impartial. But if the training data is biased the machine learning will be.
Many machine learning algorithms can’t explain how they arrived at a decision.
There is a lot of concern about how machine learning can be used for unethical purposes.
No machine learning system is 100% accurate, think about the implications of false positives and false negatives.
|
Find out more
|
This course has only touched on a few areas of machine learning.
Machine learning is a large and growing field.
This course is designed to teach you just enough to do something useful.
Machine learning is a rapidly developing field and new tools and techniques are constantly appearing.
|