Machine Learning Introduction

Machine Learning Introduction

In this blog, we will be discussing the general understanding of the machine learning and deep learning model how they work and how we can make them work in the way we want to design it to function and react to the conditions being provided to it at different times.

Basic Understanding of our M.L/ D.L Model

Machine Learning Model

Corpus is the collection of our whole data X based on which our model will calculate to give out some data Y.

When our model gives out some output Y based on the data stored in corpus as X is called Machine Learning, where we generally teach our model to give some output based on a lot of data fed by us.

Deep Learning Model

In deep learning, we still have the concept of input data (X) and output (Y), similar to machine learning. However, deep learning models are characterized by their use of neural networks, which consist of multiple layers of interconnected nodes or neurons. These neurons are organized into layers, typically including an input layer, one or more hidden layers, and an output layer.

The Input layer represents the initial data that you feed into the deep learning model. In the context of deep learning, this input data is often represented as a feature vector or a multidimensional array.

The Hidden Layers are the intermediate layers between the input and output layers. They are called "hidden" because you typically don't have direct access to their values. Each neuron in a hidden layer receives input from the previous layer and processes it through an activation function. Multiple hidden layers allow the model to learn increasingly complex representations of the data.

The Output Layer produces the final output of the deep learning model. The format of the output depends on the specific task. For example, a classification task, it might produce probabilities for different classes, while in a regression task, it could produce a numerical value.

How does it Work?

Training Data is used for training our model by which our machine learning model would be created out of the algorithm.

Mode is a Mathematical equation like for example:

y = mx + c

Then here x is the input that we provided and y is the output being produced by the machine learning model.

Here m and c are the parameters of our model which will be obtained after training our model, leading to generating these parameters. Parameters are often called Weights as well.

Algorithm is something via which we will obtain our model.

Train Data is something that will be trained over a lot of data with some errors and fixed over a vast database to make it go through the whole process of learning and applying.

Test Data is something whose training has been completed already by some calculations and now remains to just give the data or output of something that has been already tested out.

Types of Learnings

Supervised Learning is something in which we know what will be Y if X is input.

While Unsupervised Learning is something in which we won't be knowing Y.

Now Supervised and UnSupervised Learnings are classified further into:-

Supervised Learnings

Supervised Learning is something in which we know what will be Y if X is input.

  • Regression

    • There is a continuous flow of output Y on the input X.
  • Classification

    When data falls in some classification of data then that is classification type.

Unsupervised Learning

  • Clustering

    For data, if we don't know which class our data belongs to, then it is known as clustering.

    Like in Classifications we already have classes, we just have to assign values & data to those classes. But in Clustering, we may have some classes or might not have any as a result, we need to create those classes and then assign data or clusters to those classes.

Classes are some sections like Human Class and clusters are the points which makes up those classes, like we all are clusters to the Human class.

  • Association rule learning is a type of unsupervised learning technique that checks for the dependency of one data item on another data item and maps accordingly so that it can be more profitable. It tries to find some interesting relations or associations among the variables of the dataset. It is based on different rules to discover the interesting relations between variables in the database.

Re-enforcement Learning

Figuring out what actions to take based on the given current state of the environment, based on which our agent performs those actions is known as Re-enforcement Learning.

Re-enforcement Learning is a sort of hit-and-trial method in which our Agent, performs some actions based on the current state based on which it gets a reward if it does the right action & that's how we get to teach our model/ Agent to do the right thing.

Hope you get to learn some from this blog for which you came here 😄. Soon I will be releasing other blogs and building some real-time models based on supervised and unsupervised learning for facial expression readers, Face recognition and Gesture & Expression controllers for games and systems & a lot more, with all deep learning and machine learning concepts as well.

If you like my Article then please react to it and connect with me on Twitter if you are also a tech enthusiast. I would love to collaborate with people and share the experience of tech😄😄.

My Twitter Profile:

Aryan_2407

Did you find this article valuable?

Support Aryan Parashar by becoming a sponsor. Any amount is appreciated!