Ai Project Life Cycle

A typical AI project life-cycle generally includes six main steps: problem definition, data gathering, data preparation, model building, evaluation and refinements, and deployment.
Data gathering is the process of collecting relevant data for the problem.
Evaluation & Refinements involve testing the model’s performance and improving it.
Deployment means putting the model into production for real-world use.
Data cleaning is not generally listed as a separate step because it is part of the larger data preparation phase.
In practice, data cleaning is an activity done within data preparation to remove errors, handle missing values, and ensure high-quality data.
Therefore, while important, data cleaning is not counted as one of the six main steps on its own.
So, the correct answer is option (B) Data cleaning.

The Design or Build stage in an AI project involves multiple activities such as data preparation, model building, training, and testing.
This stage is inherently iterative, meaning the process is repeated many times to improve model accuracy.
Engineers and data scientists test different algorithms, tune parameters, and validate results.
It is not the first stage — that is problem definition.
It does not take a fixed time like 15 days because the duration depends on project complexity.
Also, it includes more than just data acquisition — it covers feature engineering, model training, and validation.
So, the correct statement is that the Design or Build phase is an iterative process.

An AI project starts with clearly defining the problem that needs to be solved.
This step involves understanding the business objective, setting goals, and identifying constraints.
Without a well-defined problem, it is impossible to determine what data is required or how the model will be evaluated.
Once the problem is defined, the next steps like data gathering, model construction, and deployment follow in order.
So, the first step in the AI project process is always Problem definition.

The typical AI project life-cycle consists of three key stages: Project Scoping, Design or Build Phase, and Deployment in Production.
Project scoping involves defining the problem, setting objectives, and outlining deliverables.
Design or build phase includes data collection, model development, training, and testing.
Deployment in production means integrating the AI solution into real-world use cases and monitoring its performance.
Data visualization, on the other hand, is a technique used within various stages, especially during data analysis and reporting.
It is not a standalone phase of the AI life-cycle but a supporting task to help interpret and communicate data insights.
Therefore, the correct answer is option (B) — Data visualization is not one of the three main stages.

A capstone project is indeed the final project undertaken by students at the end of their academic program.
It serves as a culminating academic and intellectual experience, demonstrating the skills and knowledge gained.
Additionally, a capstone project usually requires students to independently research a specific topic or problem.
This promotes critical thinking, problem-solving, and application of concepts in a practical way.
Therefore, both statements accurately describe what a capstone project involves.
So, the correct answer is that both Statement I and Statement II are correct.

Building the Model is a core technical stage in the AI project life-cycle.
After understanding the requirements and preparing the data, the team develops and trains the AI model.
This involves selecting appropriate algorithms, setting parameters, and feeding clean data to the model.
The model learns patterns, relationships, and trends from the data through training.
It may go through multiple iterations to improve accuracy, reduce errors, and prevent overfitting.
Techniques like cross-validation, hyperparameter tuning, and performance metrics are applied to build an effective model.
Developers also compare different models to choose the best one.
A well-built model is essential for making reliable predictions and solving the defined problem effectively.
Once trained and tested, the model moves forward for evaluation and deployment.

Requirements Analysis is one of the first and most important stages in any AI project life-cycle.
In this stage, the project team works closely with stakeholders to clearly define the problem that needs to be solved.
This includes identifying the goals, scope, limitations, and success criteria for the AI project.
Key tasks in this phase are gathering detailed business requirements, understanding data availability, and setting performance expectations.
Proper requirements analysis helps avoid misunderstandings and sets a strong foundation for the project.
It also ensures that the final AI solution aligns with the actual needs of the business or user.
Missing or unclear requirements can lead to wasted time, resources, and a failed project.
So, this stage acts like a blueprint for the entire AI project life-cycle.

The Requirement Analysis stage is the initial phase of any AI project cycle.
This stage focuses on understanding what the project aims to achieve and planning how to achieve it.
It involves gathering requirements from stakeholders, defining the scope, and setting clear objectives.
This stage also includes analyzing constraints, risks, and the motivational aspects that drive the project forward.
Design, testing, and deployment come later and depend on clear requirements being defined first.
Without solid requirement analysis, the project can lose direction or fail to meet its goals.
So, the correct answer is option (D) — Requirement Analysis.

The scoping phase of an AI project involves planning and defining what the project aims to achieve.
Key activities include clarifying business goals, setting success criteria, and making sure all stakeholders have aligned expectations.
Evaluating ROI can also be part of scoping to check if the project is worth pursuing.
However, cleaning and preparing the data is not part of scoping — it happens later during the data preparation phase.
Data cleaning involves removing errors, handling missing values, and transforming raw data into usable form.
Scoping is about answering “what” and “why”; data cleaning answers “how” during the build phase.
So, the correct answer is option (C) — Cleaning and preparing the data is not a scoping activity.

Step 1: Define preprocessing in computer vision.
Preprocessing involves preparing raw images for further analysis by improving their quality and removing noise.

Step 2: Role of preprocessing.
It includes tasks like resizing, normalization, filtering, and noise reduction to make images suitable for processing.

Step 3: Importance of preprocessing.
It enhances image clarity and helps algorithms perform more accurately and efficiently.

Step 4: Define high-level processing.
High-level processing involves interpreting the image, such as object detection, recognition, and decision-making.

Step 5: Difference between both.
Preprocessing focuses on improving image quality, while high-level processing focuses on extracting meaning and insights from the image.

Step 1: Understanding the problem.
The images contain random dots and distortions caused by low lighting conditions, which are referred to as noise.

Step 2: Identifying the technique.
• (A) Cropping: Removes parts of an image. • (B) Noise Reduction: Correct. Removes unwanted distortions and improves image clarity. • (C) Resizing: Changes image dimensions. • (D) Normalization: Adjusts pixel intensity values.
Step 3: Conclusion.
Thus, the technique used to remove distortions is Noise Reduction.
Final Answer: Noise Reduction.