AI PROJECT CYCLE

DATA MODELLING
Topics to be Data Modelling

covered Categories of AI Models

Categories of Learning Based AI Models

 DATA MODELLING
Data Modelling is the phase during which the AI model is trained using the collected data repeatedly until it
produces the desired results.
 Categories of AI Models
Rule Based
AI Models Machine Learning
Learning
Based
Deep Learning
 Rule Based AI Models
Rule Based AI Models:- They are developed using the algorithms having pre-defined labels,
rules, patterns and relationships.
Drawbacks of Rule Based AI Models:-
i. Consumes a lot of time:- Creating all possible rules for a system requires a lot of time. Bigger and more
complex system is, more time consuming it becomes.
ii. Limited adaptability and learning capacity.
iii. Static in nature and not scalable.
 Learning Based AI Models
Learning Based AI Models are trained to learn by inputting them tons of data. Here, there are no
patterns, rules and relationships predefined by developer, rather machine learns with each new input and
comes up with its own algorithm.

This style is appropriate when data to be processed is unlabeled and too random too fit into the frame of
common rules and patterns.

Rather, random data is repeatedly fed into the machine and it analyses each input and tries to figure out
patterns and trends out of the input collectively.
 RULE BASED VS LEARNING BASED
Key Rule Based Learning Based
Rule Based Approach Refers to the AI Whereas in Learning based approach, the
modelling where the relationship or patterns in relationship or patterns in data are not
data are defined by the developer. The machine defined by the developer. In this
follows the rules or instructions approach, random data is fed to the
Approach
mentioned by the developer, and performs its machine and it is left to the machine to
task accordingly. figure out patterns and trends out of it.

The machine once trained, doesn’t take into This type of approach facilitates the
Mutability consideration any changes made in the original acquisition of new knowledge.
dataset.
 Machine Learning
Machine Learning (ML) is a branch of AI that enables machines to progressively learn
and improve at tasks by the use of data and without being programmed explicitly. The
recommendation system on music and video streaming services are example of ML.

Supervised
Learning

LEARNING
MACHINE
Unsupervised
Learning

Reinforcement
Learning
 Supervised Learning
Supervised Learning is a learning approach of machine when the machines with the help of algorithm
learns from a labelled dataset and is latter tested with unlabeled dataset whose answers are pre-known to
evaluate the accuracy on training data.

Supervised
Learning

Regression Classification
 Regression
Regression is a mathematical approach to find
relationship between two or more variables. It works
with continuous data. For example, if you wish to
predict your next salary, then you would put in the
data of your previous salary, any increments, etc., and
would train the model. Here, the data which has been
fed to the machine is continuous.
Examples:-
➢ Price of House prediction
➢ Temperature of the city
➢ Sales Prediction
 Classification
In classification, data is categorized under different labels according to some
parameters given in input and then the labels are predicted for the data. This
model works on discrete dataset which means the data need not be continuous.

For example, we have 3 coins of different denomination which are labelled

according to their weight then the model would look for the labelled features
for predicting the output.

Other Examples:-
➢ Is this email spam or not?
➢ Is this Social Media Post positive or negative?
➢ Genre of song/movie.
 Regression VS Classification
Key Regression AI Model Classification
In Regression AI Model, we try to find the In Classification AI Model, we try to find the best
best-fit line which can represent the overall possible decision boundary which can separate the
Definition trend in the data. two classes with the maximum possible
separation.

Output It is used with continuous data. It is used with discrete data.

Problems like House Price Prediction, Problems like Spam Email Classification, Disease
Rainfall Prediction like problems are solved prediction like problems are solved using
Examples
using Regression AI Model. Classification AI Model.
 Unsupervised Learning
An Unsupervised Learning Model works on unlabeled dataset. This means that the data which is fed to
the machine is random and there is a possibility that the person who is training the model does not have any information
regarding it. The unsupervised learning models are used to identify relationships, patterns and trends out of the data which
is fed into it. It helps the user in understanding what the data is about and what are the major features identified by the
machine in it.
Unsupervised
Learning

Dimensionality
Clustering
Reduction
 Clustering
This is a Machine learning approach where the machine generates its own
rules or algorithms to differentiate amongst the given dataset to achieve the
pre-decided goal. The data fed to such a model is usually unlabelled or
random and thus the developer feeds in the data directly into the machine
and instructs it to build its own algorithm. The machine then finds out

patterns or trends out of the training dataset and

clusters the ones which follow the same pattern. The output rules might be
very different to what was expected as the machine has its own way of
recognising patterns.

Examples
➢ Pattern Recognition
➢ Document Analysis
➢ Identifying Fake News
 Clustering VS Classification
Key Clustering AI Model Classification AI Model
Clustering is an unsupervised learning approach. Classification is a supervised learning approach.
Approach

It groups the instances based on how similar they It is a process where the input instances are
are, without using class labels. classified based on their respective class labels.
What does it do?

It is not needed to train and test the dataset. It has labels, hence there is a need to train and test
Training and
the dataset to verify the model.
testing
 Reinforcement Learning
In Reinforcement Learning approach, the AI Model (the algorithm, also called as agent) iteratively
attempts to accomplish a particular goal, or improve performance on a specific-task, in the best
possible way known to it.
In this learning approach the agent learns automatically by using hit and trial methods or through
its own experience using rewards and penalties. Each action performed by the agent give rewards for
correct move and for wrong move it generates negative feedback.

The main element of RL:-

The agent or the learner.
The environment the agent interacts with.
The policy that the agent follows to take actions.
The reward signal that the agent receives after taking actions.
 Reinforcement Learning…Contd
Practical Applications of RL;-

➢ Training Robots to work in hazardous situation and terrains.

➢ Training Robots to perform dangerous and difficult tasks in various industries.

➢ In gaming for learning ways to better and better game play and strategies.
THANK YOU