Skip to main content

Types of Machine Learning Systems

 There are so many different types of Machine Learning systems that it is useful to

classify them in broad categories based on:

• Whether or not they are trained with human supervision (supervised, unsuper‐

vised, semisupervised, and Reinforcement Learning)

• Whether or not they can learn incrementally on the fly (online versus batch

learning)

• Whether they work by simply comparing new data points to known data points,

or instead detect patterns in the training data and build a predictive model, much

like scientists do (instance-based versus model-based learning)

These criteria are not exclusive; you can combine them in any way you like. For

example, a state-of-the-art spam filter may learn on the fly using a deep neural network model trained using examples of spam and ham; this makes it an online, model-

based, supervised learning system.

Let’s look at each of these criteria a bit more closely.

Supervised/Unsupervised Learning

Machine Learning systems can be classified according to the amount and type of

supervision they get during training. There are four major categories: supervised

learning, unsupervised learning, semisupervised learning, and Reinforcement Learning.

Supervised learning

In supervised learning, the training data you feed to the algorithm includes the desired

solutions, called labels.


A typical supervised learning task is classification. The spam filter is a good example

of this: it is trained with many example emails along with their class (spam or ham),

and it must learn how to classify new emails.

Another typical task is to predict a target numeric value, such as the price of a car,

given a set of features (mileage, age, brand, etc.) called predictors. This sort of task is

called regression (Figure 1-6). 1 To train the system, you need to give it many examples

of cars, including both their predictors and their labels (i.e., their prices).


Note that some regression algorithms can be used for classification as well, and vice

versa. For example, Logistic Regression is commonly used for classification, as it can

output a value that corresponds to the probability of belonging to a given class (e.g.,

20% chance of being spam).

Here are some of the most important supervised learning algorithms (covered in this

book):

• k-Nearest Neighbors

• Linear Regression

• Logistic Regression

• Support Vector Machines (SVMs)

• Decision Trees and Random Forests

• Neural networks 2

Unsupervised learning

In unsupervised learning, as you might guess, the training data is unlabeled

(Figure 1-7). The system tries to learn without a teacher.


Here are some of the most important unsupervised learning algorithms (we will

cover dimensionality reduction in Chapter 8):

• Clustering

— k-Means

— Hierarchical Cluster Analysis (HCA)

— Expectation Maximization

• Visualization and dimensionality reduction

— Principal Component Analysis (PCA)

— Kernel PCA

— Locally-Linear Embedding (LLE)

— t-distributed Stochastic Neighbor Embedding (t-SNE)

• Association rule learning

— Apriori

— Eclat

For example, say you have a lot of data about your blog’s visitors. You may want to

run a clustering algorithm to try to detect groups of similar visitors (Figure 1-8). At

no point do you tell the algorithm which group a visitor belongs to: it finds those

connections without your help. For example, it might notice that 40% of your visitors

are males who love comic books and generally read your blog in the evening, while

20% are young sci-fi lovers who visit during the weekends, and so on. If you use a

hierarchical clustering algorithm, it may also subdivide each group into smaller

groups. This may help you target your posts for each group. 


Visualization algorithms are also good examples of unsupervised learning algorithms:

you feed them a lot of complex and unlabeled data, and they output a 2D or 3D rep‐

resentation of your data that can easily be plotted (Figure 1-9). These algorithms try

to preserve as much structure as they can (e.g., trying to keep separate clusters in the

input space from overlapping in the visualization), so you can understand how the

data is organized and perhaps identify unsuspected patterns.

Comments

Post a Comment

Popular posts from this blog

Possible Limitations of AI-Based Bots

 The examples above already show the present-day potential of AI-based bots. At present, these systems are still in an early stage and still have certain limitations and potentials for optimisation. Twitter Bot Tay by Microsoft Most bots at present are reactive service bots. Engagement bots that actively interact with the users as market and brand ambassadors go one step further. The most famous example here is the chatbot Tay by Microsoft. Microsoft removed Tay from the web apologetically within one day. The example shows that the uncontrolled training of bots by the community can lead to fatal consequences. AI systems still have to learn ethical standards. It thus becomes apparent that even bots require a kind of guideline. Like a journalist has to observe editorial guidelines, bots have to observe certain standards. The next generation of AI-based bots must control and create the possible room for communication. IBM Watson has been able to celebrate quite a few respectable resul...
Post a Comment
Read more

What is Machine Learning

 The term machine learning (ML) as a part of artificial intelligence is ubiq- uitous nowadays. The term is used for a wide number of various appli- cations and methods that deal with the “generation of knowledge from experience”. The well-known US computer scientist Tom Mitchell defines machine learning as follows: A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E (Mitchell 1997). An illustrative example of this would be a chess computer program that improves its performance (P) in playing chess (the task T) by experience (E), by playing as many games as possible (even against itself ) and analysing them (Mitchell 1997). Machine learning is not a fundamentally new approach for machines to generate “knowledge” from experience. Machine learning technology was used to filter out junk e-mails a long time ago. Whilst spam filters that tack- ...
Post a Comment
Read more

What is Data Protection and Data Integrity

 As a matter of principle, when it comes to data protection, a differentiation must be made between personal data and data involving companies. As soon as inferences can be made to a specific individual and single data levels are being worked at, a moment has to be taken to consider: What is being processed? Is there already a business relationship? Which permissions or legal consent elements are at hand? Customer data may not be collected without permission and may also not be resold. Anybody who acts carelessly here can quickly render themselves liable to prosecution. In principle, the following applies however: Almost anything is possible with the customer’s consent. This is the reason why Facebook can act with the data to such an extent, because consent has been given, even if only few users have probably fully read and understood the Terms of Use. Likewise, a relatively far-reaching data processing in the scope of an ongoing customer relationship under the motto “for our own p...
Post a Comment
Read more