# Putting Supervised Learning into Practice

**Introduction**

Supervised learning is a machine learning technique that involves feeding your algorithm with labeled examples to train it. The label is the information (classification) you want to predict. For example, if you’re predicting if a cat is on a given picture or not, the label would be ‘cat’. A common type of supervised learning algorithm is Support Vector Machines (SVMs). Another common type of supervised learning algorithm is logistic regression.

**Supervised learning is a machine learning technique that involves feeding your algorithm with labeled examples to train it.**

Supervised learning is a machine learning technique that involves feeding your algorithm with labeled examples to train it. The label is the information (classification) you want to predict, and it can be anything from what type of plant is in front of you or whether or not there’s cancer in your body.

The example below shows how supervised learning works:

- First, we take our training data set and split it into two parts: one for training our model and another for testing its performance on new data later on. In this case, we’ll call these two sets “train” and “test.”
- Then we use Kaggle’s random forest algorithm (a supervised learning method) to train our model using both sets of data together as input parameters–the Xs–and labels as outputs–the Ys.

**The label is the information (classification) you want to predict. For example, if you’re predicting if a cat is on a given picture or not, the label would be ‘cat’.**

The label is the information (classification) you want to predict. For example, if you’re predicting if a cat is on a given picture or not, the label would be “cat”. If you’re trying to predict whether someone will develop heart disease or not in their lifetime and your dataset contains people who have already developed heart disease, then “Yes” would be used as the label for those cases and “No” for all other cases.

The most common supervised learning algorithms are:

- Logistic Regression – This algorithm can be used when there are two classes of data being predicted with categorical values such as ‘yes/no’ or ‘male/female’. The output from this algorithm will result in an equation that predicts whether something falls within one category or another based on its input variables (x1…xn).

**A common type of supervised learning algorithm is Support Vector Machines (SVMs).**

One common example of supervised learning is Support Vector Machines (SVMs). SVMs are a way to find a line that separates the data into two different classes. In this case, you want your algorithm to separate “good” customers from “bad” ones. Using this method, we can train our algorithm on each customer’s transaction history and then use it to predict whether or not they will default on their loan payment in the future.

If you’ve studied linear algebra in school or college before, then you probably know about vectors–these are just lists of numbers that have an associated direction and magnitude. In other words: if I give someone a vector like [1 2 3] , they know it represents three points in space–they could even draw these points on paper! However, there are many different types of vectors besides just simple lists of numbers like these (for example: matrices). And by using certain operations on these kinds of vectors we can create even more complex objects called tensors which allow us do cool things like multiply two matrices together without having to convert them into single-dimensional arrays first…

**Another common type of supervised learning algorithm is logistic regression.**

Logistic regression is a statistical method for modeling the probability of an event. It’s used in many different types of applications, including:

- Predicting whether someone will have a disease based on medical data;
- Classifying emails as spam or legitimate; and even more mundane tasks like finding out which products customers buy most often based on their shopping history.

**One way to get started with supervised learning is by using traditional algorithms from the field of statistics.**

One way to get started with supervised learning is by using traditional algorithms from the field of statistics. The most common approach is to use a dataset to train the algorithm, then validate it against another dataset that wasn’t used during training. Finally, you can use this third dataset as a way of seeing how well your algorithm generalizes to unseen data.

The first step in this process is choosing an appropriate machine learning algorithm for your problem (we’ll get into those later). Once you’ve picked one out and put it into practice on your training set, you should have some idea whether or not its results are reliable enough for further testing–but which tests should you perform?

**Conclusion**

In this article, we’ve covered some of the most popular supervised learning algorithms and how they work. There are many more ways to use supervised learning in practice, but this should give you a good starting point. If you want to learn more about how other types of machine learning work (unsupervised or reinforcement), check out our other posts on those topics!