AI PROJECT CYCLE

It involves 5 steps

1) Problem Scoping

2)Data Acquisation

3)Data Exploration

4)Modelling

5)Evaluation

Problem Scoping

It is the first step of the Project Cycle here we identify the problem and get a clear understanding of it
and ways to fix it

Data Acquisition

It is the second step of the project cycle here we obtain and collect data from different authentic and
reliable sources. This is crucial as data will come in large quantities. Previous data is called TRAINING
DATA and the prediction data is called TESTING DATA.

Data Exploration

It is the third step of the project cycle, it is the process of analyzing and understanding the data before
the AI model. In this stage we explore different patterns. We use summary statistics, Visulatizations etc.
This step helps in finding any missing data and makes it ready for AI model. Examples of some open
sourced websites are :data.gov.in, Indian.gov.in.

Modelling

It is the fourth step of the project cycle, after exploring patterns, we have to choose a model you would
build to achieve the goal. You can search online for various models.

Evaluation

This is the final stage of the project cycle, Evaluation in the AI project cycle is the stage where
you test how well the AI model is performing after it has been trained. In this phase, the model's
predictions are compared against actual results using a separate set of data (called the test data).
The goal is to measure the model’s accuracy. During evaluation, different metrics like F1-score
are used to see how good the model is at making correct predictions or decisions. The model is
tested with the help of Testing Data
SDGS GOALS

1. No Poverty
2. Zero Hunger
3. Good Health and Well-being
4. Quality Education
5. Gender Equality
6. Clean Water and Sanitation
7. Affordable and Clean Energy
8. Decent Work and Economic Growth
9. Industry, Innovation, and Infrastructure
10. Reduced Inequality
11. Sustainable Cities and Communities
12. Responsible Consumption and Production
13. Climate Action
14. Life Below Water
15. Life on Land
16. Peace, Justice, and Strong Institutions
17. Partnerships for the Goals

4Ws Problem Canvas

WHO,WHAT,WHERE AND WHY.

WHO: Helps in analysing the people getting affected directly or indirectly. Under this there are
stakeholders they are the one who face the problems

WHAT: Here we find out the nature of the problem and what the problem is. Here we also gather
evidence to prove that problem actually exists. Ex; Newspaper articles

WHERE: Here we have to find out the location, and where does the problem arise from

WHY: Explains the purpose behind the idea or project. And that why does the solution matter to the
community.
AI MODELS : IT CAN BE CLASSIFIED AS

1. LEARNING BASED – ML,DL ( MACHINE LEARNING ), (DEEP LEARNING).

RULE BASED

Rule-based data refers to a type of data system or methodology where decisions or actions are
determined based on a set of predefined rules. These rules are typically created by experts and are
used to guide the processing and interpretation of data.

DRAWBACK OF RULEBASED DATA

• A drawback/feature for this approach is that the learning is static.

• If the machine is trained once you cant change it.

It is when the machine learns by itself. Learning-based data is when models or systems improve
their performance over time based on data and experience. Learning Based data doesn’t have
predefined rules inbuilt in it. This is commonly associated with machine learning and artificial
intelligence.

They have 3 types:

Supervised Learning

• In a supervised learning model, the dataset which is fed to the machine is labelled. A label is
some information which can be used as a tag for data.

2 types of Supervised Learning

Classification: Where data is classified according to the labels. Here the data need not be
continuous.

Regression : These modes work on continuous data. Predicts numerical values.

• Works on unlabelled 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.

2 TYPES:

Clustering

It is a type of unsupervised learning where it can cluster unknown data according to the
patterns.

Dimensionality Reduction

Dimensionality reduction helps simplify the data, making it easier to visualize, analyze, and use
in machine learning models. To reduce the dimensions and still be able to make sense out of the
data, we use Dimensionality Reduction.

Reinforcement Learning

Reinforcement Learning (RL) is a type of machine learning where an agent learns to make
decisions by performing actions in an environment to maximize a cumulative reward.

NEURAL NETWORKS

• Neural network are modelled after the neurons in our brain. The key advantage of neural
networks are that they are able to extract data features automatically without needing the input
 of the programmer. A neural uses algorithms to perform certain tasks. It is a fast and efficient
way to solve problems for which the dataset is very large, such as in images.

• ( NOTE: LARGER NEURAL NETWORK TEND TO PERFORM WITH LARGER AMOUNTS OF DATA. )

• A Neural Network is divided into multiple layers and each layer is further divided into several
blocks called nodes.

• FIRST LAYER : INPUT LAYER

• The job of an input layer is to acquire data and feed it to the Neural Network. No
processing occurs at the input layer.

SECOND LAYER : HIDDEN LAYER

• . Hidden layers are the layers in which the whole processing occurs. Their name

essentially means that these layers are hidden and are not visible to the user. The last hidden
layer passes the final processed data to the output layer which then gives it to the user as the
final output.

THIRD LAYER : OUTPUT LAYER

It is designed specifically to be used in the part of a computer or app that people

interact with directly.( It is meant for user-interface).

SOME OF THE FEATURES OF A NEURAL NETWORK ARE LISTED BELOW:

. They are able to automatically extract features without input from the programmer

. It useful when solving for large data set

.They are modelled on the humain brain and network system

3 CATOGARIES OF DATA

• Training Data : Old or Previous Data.

• Validating data : Also called secondary data set. This data is used to check if the model is
correctly identifying the data for making predictions.

• Testing data : Predicting or Prediction Data

Data warehousing
• Data is always collected in bulk from various sources using various formats. This is called data
warehousing.