AI Project Cycle Class 9 Notes

These AI Class 9 Notes Chapter 2 AI Project Cycle Class 9 Notes simplify complex AI concepts for easy understanding.

Class 9 AI Project Cycle Notes

Project Class 9 Notes

A project is a series of tasks that need to be completed to reach a specific outcome. A project can also be defined as a set of inputs and outputs required to achieve a particular goal. Projects can range from simple to complex and can be managed by one person or a hundred.

Project Life Cycle

A project life cycle is a series of phases that a project goes through from initiation to completion. These phases typically include initiation, planning, execution, monitoring and controlling, and closure. Each phase has specific tasks, deliverables, and goals aimed at achieving project objectives.

AI Project Cycle Class 9 Notes

The AI project cycle is a step-by-step process that uses scientific methods to solve problems and draw conclusions. The AI project cycle is a structured roadmap for developing and deploying artificial intelligence projects to solve real-world problems. It guides organizations and individuals through a structured process that includes problem scoping, data acquisition, data exploration, modelling, and evaluation. This systematic approach helps define objectives, collect relevant data, uncover hidden insights, create models for AI systems, and assess their performance. By following the AI project cycle, stakeholders can enhance the efficiency and effectiveness of their AI initiatives.

The phases of the AI project cycle include

1. Problem Scoping
2. Data Acquisition
3. Data Exploration
4. Modelling
5. Evaluation
6. Deployment

Problem Scoping

• Problem Scoping is the first phase of the AI project cycle. In this stage of project development, problems will be identified. It is then followed by designing, developing, or building, and finally testing the project.
• Problem Scoping means selecting a problem which we might want to solve using our AI knowledge. The process of identifying the aim and scope of the problem that you wish to solve with the help of your project. This is “Problem Scoping”.
• The problem scoping refers to the identification of a problem and the vision to solve it. AI project cycle everything will be failed if problem scoping is failed or without appropriaté problem scoping. Incorrect problem scoping also leads to failure of the project as well.

The 4Ws Problem Canvas

Problem scoping is the process of pinpointing a particular issue or opportunity that can be tackled using artificial intelligence. During this phase, we not only identify the problem but also set specific objectives, goals, and criteria for success. However, scoping a problem is no simple task. It requires a deep understanding of the issue so that we can work effectively and solve problem-solving.

To achieve this, we rely on an approach called the 4Ws problem canvas, which helps us gain a clearer and more defined understanding of the problem we’re dealing with.

The 4Ws Problem canvas helps you in identifying the key elements related to the problem. Let us go through each of the blocks one by one.

Who This identifies all the stakeholders involved. Stakeholders are anyqne impacted by the problem, directly or indirectly. This could be users, customers, employees, managers, or even external parties. Understanding who’s affected helps tailor solutions.

What This focuses on the core problem itself. What exactly is the issue? Here, you gather evidence to confirm the problem exists and define its scope. Be specific! A vague “low sales” problem becomes more actionable when you pinpoint a drop in online sales or a decline in sales of a specific product line.

Where This considers the context and location of the problem. Does it occur everywhere, or in a specific department, region, or situation? Understanding the environment can help pinpoint root causes and guide potential solutions. For example, high product returns might be due to poor packaging during shipping, or unclear instructions included with the product.

Why This explores the importance of solving the problem. Why is it worth addressing? What are the negative consequences if left unaddressed? Highlighting the impact (e.g. lost revenue, customer dissatisfaction) strengthens the case for finding a solution.

These three canvases now become the base of why you want to solve this problem. Thus, in the “Why” canvas, think about the benefits which the stakeholders would get from the solution and how would it benefit them as well as the society.

The Problem Statement Template When the above 4Ws are filled, you need to prepare a summary of these 4 Ws.

This summary is known as the problem statement template. This template explains all the key points in a single template. So, if the same problem arises in the future this statement helps to resolve it easily. We can take a look at the Problem Statement Template and understand the key elements of it.

Data Acquisition

As we move ahead in the AI project cycle, we come across the second stage which is data acquisition. As the term clearly mentions, this stage is about acquiring data for the project. While developing an AI system for predictive purposes, it’s essential to begin by training it with relevant data.

Data acquisition consists of two words

• Data Data refers to the raw facts, figures, information, or statistics. Data can be in the format of the text, video, images, audio, and so on and it can be collected from various source like interest, journals, newspapers and so on.
• Acquisition Acquisition refers to acquiring data for the project.
• So, Data acquisition means acquiring data needed to solve the problem.
• Suppose you want to build a system that predicts how much an employee will earn in the future based on what they’ve earned in the past.
• To do this, you feed the system the historical salary information. We call this historical salary data “training data”, while the dataset used for future salary predictions is called testing data.
• Here, the specific kinds of information you want to collect are called data features. In our previous example, these data features could be things like the employee’s salary amount, the percentage of salary increase they received, the time between salary raises, any bonuses they’ve earned, and more.

There are various methods to gather this data, such as follows

1. Surveys
2. Web scraping
3. Sensors
4. Cameras
5. Observations
6. API (Application Programming Interface)

One of the most dependable and trustworthy sources for data is government-hosted open-access websites. Examples of such government portals include data.gov.in and india.gov.in.

Data Features

• Data features refer to the type of data you want to collect. In our previous example of future salary prediction, data features would be salary amount, increment percentage, increment period, bonus, etc.
• In AI, data features are used to describe the input data that we feed into a machine learning model. These features help the model understand and make predictions.
• Example Suppose you want to predict the price of a house. The features could include the number of bedrooms, the size of the house, the location, and the age of the house.

Type of data you want to collect. Here two terms are associated with this

1. Training Data Training data is a set of examples used to teach an AI model how to make predictions or perform a task. It’s like giving practice problems to a student to help them learn. The model learns from the patterns in the training data and adjusts its internal parameters accordingly.

Example Imagine you’re teaching a robot to recognize different animals. You show it pictures of cats, dogs, and birds along with labels indicating what each animal is. The robot learns from these examples so that when you show it a new picture, it can correctly identify the animal.

2. Testing Data Testing data is another set of examples that we use to evaluate how well our AI model performs after it has been trained. It’s like giving a student a test after they’ve studied to see how well they’ve learned the material.
Testing data is separate from the training data to ensure that the model can generalize well to new, unseen examples.

Example Going back to the animal recognition example, after training the robot with lots of pictures of animals, you give it a new set of pictures that it hasn’t seen before. This testing set helps you assess if the robot can correctly identify animals it hasn’t encountered during training.

The testing data quality is essential for the machine learning model’s performance, The data should represent the real world and be large enough to evaluate the model thoroughly.
Note The training data and the test data are not different, they are usually divided from the main dataset in 80%-20%.

System Maps

A system map shows the components and boundaries of a system and the components of the environment at a specific point in time. With the help of System Maps, one can easily define a relationship amongst different elements which come under a system. Relating this concept to our module, the Goal of our project becomes a system whose elements are the data features mentioned above. Any change in these elements changes the system outcome too. For example, if a person received 200 % increment in a month, then this change in his salary would affect the prediction of his future salary. The more the increment presently, the more salary in future is what the system would predict.

So how do we create a systems map? Here is a step-by-step guide to make a systems map.
The Steps

1. Developing a systems map relies on the thoroughness of knowledge about the problem. That is, when you have more information about the problem, your systems maps can be more detailed.
2. The best way to start once you enough information, is to write statements about how X leads to Y and under what conditions.
3. The next step is to separate information that belongs together and that does not.
4. Separate the contextual factors.
5. Establish connections between the major chunks that go together.
6. Gradually follow the chronological order to develop a narrative.
7. Create a systems diagram.
8. Refine the diagram till it accurately captures the story and provides information that is easier to understand.
9. Make sure you refine the diagram in such a manner that the connection between the main themes can be made explicit.

Data Exploration

Data exploration is the third stage of AI project cycle. Data exploration is the process of arranging the gathered data uniformly for a better understanding. Data can be arranged in the form of a table, plotting a chart or making database. It is the first step of data analysis, after data acquisition, to explore and visualize data to uncover insights from the raw data. The extracted datá then has to be studied using various techniques for understanding the structure of the data and identifying the trends, patterns, and inter-relationships within the data.

Data is a complicated, often just a bunch of numbers. But to make sense of it, we need to find the hidden patterns. That’s where data visualization comes in. It’s all about turning those numbers into pictures that are easy for people to understand. This allows you to include:

Spotting Trends and Patterns When you look at a chart or graph, you can quickly see if something is going up, down, or staying the same. That helps you spot trends and patterns in the data.

Choosing the Right Tools Data visualization helps you decide which method or model is best for further analysis. It’s like picking the right tool for the job.

Sharing Insights Once you’ve figured things out, you can show your findings to others. A picture is worth a thousand words, as they say. Visualizing data makes it easier to explain your insights to others.

To visualize data, you can use various types of visuals like bar graphs, histograms, line charts, and pie charts. These visuals make the data more approachable and understandable.

Data Visualization

Data visualization is the graphical representation of information and data in a pictorial or graphical format. Data visualization tools provide an accessible way to see and understand trends, patterns in data, and outliers. Data visualization tools and technologies are essential to analyzing massive amounts of information and making data-driven decisions.

The concept of using pictures is to understand data that has been used for centuries. General types of data visualization are Charts, Tables, Graphs, Maps, and Dashboards.

Advantages of Data Visualization

• A better understanding of data
• Provides insights into data
• Allows user interaction
• Provide real-time analysis
• Help to make decisions
• Reduces complexity of data
• Provides the relationships and patterns contained within data
• Define a strategy for your data model
• Provides an effective way of communication among users

Till now we learned about problem scoping and data acquisition. Now we have set your goal for your AI project and found ways to acquire data. When we acquired data the main problem with data is – the data is very complex. Because it’s having numbers. To make use of these numbers user need a specific pattern to understand the data.

For example, if we are going to reading a book. We went to library and selected a book. The first thing you try to do is, just turning the pages and take a review and then select a book of your choice. Similarly, when we are working with data or going to analyze data we need to use data visualization.

Data Visualization Techniques

Here are some important data visualization techniques to know

Pie Chart Pie charts are one of the most common and basic data visualization techniques, used across a wide range of applications. Pie charts are ideal for illustrating proportions, or part-to-whole comparisons.

Bar Churt The classic bar chart, or bar graph, is another common and easy-to-use method of data visualization. In this fype of visualization, one axis of the chart shows the categories being compared, and the other, a measured value. The length of the bar indicates how each group measures according to the value.

Histogram Histograms are especially useful for showing the frequency of a particular occurrence. For example, if you’d like to show how many clicks your website received each day over the last week, you can use a histogram. From this visualization, you can quickly determine which days your website saw the greatest and fewest number of clicks.

Scatter Plot An XY (Scatter) chart either shows the relationships among the numeric values in several data series or plots two groups of numbers as one series of XY coordinates.
Additionally, the closer the data points are grouped together, the stronger the correlation or tends to be.

Lime Plot Line plots are drawn by joining straight lines connecting data points where x-axis and y-axis values intersect. In Matplotlib, the plot() function represents this type of graph.

Timeline Timelines are the most effective way to visualize a sequence of events in chronological order. They’re typically linear, with key events outlined along the axis. Timelines are used to communicate time-related information and display historical data.

Area Plot The area plots spread across certain areas with bumps and drops (highs and lows) and are also known as stack plots.

In Matplotlib, the stackplot() function represents it.

Modelling

Modelling is the process in which different models based on the visualized data can be created and even checked for the advantages and disadvantages of the model.

AI Model

An AI model is a program or algorithm that utilizes a set of data that enables it to recognize certain patterns. This allows it to reach a conclusion or make a prediction when provided with sufficient information. There are 2 Approaches to make a Machine Learning Model

(i) Learning Based Approach Learning based approach is based on machine learning experience with the data fed.
The different types of learning based approaches are
Supervised Learning The data that you have collected here is labelled and so you know what input needs to be mapped to what output. This helps you correct your algorithm if it makes a mistake in giving you the answer. Supervised Learning is used in classifying mail as spam.

Unsupervised Learning The data collected here has no labels and you are unsure about the outputs. So, you model your algorithm such that it can understand patterns from the data and output the required answer. You do not interfere when the algorithm learns.

Reinforcement Learning There is no data in this kind of learning, you model the algorithm such that it interacts with the environment and if the algorithm does a decent job, you reward it, else you punish the algorithm. (Reward or Penalty Policy). With continuous interactions and learning, it goes from being bad to being the best that it can for the problem assigned to it.

(ii) Rule Based Approach A rule-based system uses rules as the knowledge representation. These rules are coded into the system in the form of if-then-else statements.

The main idea of a rule-based system is to capture the knowledge of a human expert in a specialized domain and embed it within a computer system. A rule-based system is like a human being born with fixed knowledge. The knowledge of that human being doesn’t change over time.

This implies that, when this human being encounters a problem for which no rules have been designed, then this human gets stuck and so won’t be able to solve the problem. In a sense, the human being doesn’t even understand the problem.

Evaluation

Evaluation is a process that critically examines a program. It involves collecting and analyzing information about a program’s activities, characteristics, and outcomes. Its purpose is to make judgments about a program, to improve its effectiveness, and/or to inform programming decisions.
So, Evaluation is basically to check the performance of your AI Model. This is done by mainly two thing “Prediction” and “Reality”.

• Prediction The output given by the machine after training and texting the data is known as a prediction.
• Reality This refers to the actual situation or event that the model is trying to predict. It’s the ground truth, the data point we compare the prediction to see how well the model performed.

Let’s understand this with a real-world scenario Imagine an AI model designed to predict flight delays. Here’s how prediction and reality play out in its evaluation:

• Prediction Based on historical data and current conditions, the model predicts a 20% chance of a delay for a specific flight.
• Reality The flight actually departs on time (no delay).

Here, we have two possible outcomes

• True Negative The model predicted no delay (negative), and reality confirms no delay (true). This is a successful prediction.
• False Positive The model predicted a delay (positive), but in reality, there was no delay (negative). This is an incorrect prediction.

Confusion Matrix

The comparison between the results of Prediction and reality is called the Confusion Matrix.

Prediction and Reality can be easily mapped together with the help of this Confusion Matrix.

Four Common AI Evaluation Methods

Accuracy This is a basic metric that measur ss the proportion of correct predictions made by the model. In the recommender system example, accuracy could be the percentage of users who bought recommended products.

Precision This metric focuses on the positive predictions and calculates the proportion of truly relevant items among those the model recommended. For the recommender system, precision would be the percentage of recommended products that User A actually bought.

Recall This metric considers all the relevant items and calculates the proportion of those that the model actually recommended. In the e-commerce scenario, recall would be the percentage of athletic socks User A was interested in (whether purchased or not) that the system recommended.

F1-Score This metric combines precision and recall into a single score, providing a balance between the two. It’s useful when both precision and recall are important.

Deployment

Deployment is the method by which you integrate a machine learning model into an existing production environment to make practical business decisions based on data.

Deployment is the final stage where an AI model transitions from development to real-world use. It’s like taking your creation out of the workshop and putting it to work.

Key steps in deployment involve

• Testing and validating the model to ensure it performs well in real-world situations, just like giving your creation a final test run.
• Integration with existing systems, which is like connecting your AI model to the tools and infrastructure it needs to function.
• Monitoring and maintenance, which means keeping an eye on your model’s performance and making adjustments as needed, similar to how you might fine-tune a machine over time.

Examples of deployed AI showcase the power of AI in various fields, from self-driving cars navigating the roads to medical diagnosis systems assisting doctors to chatbots providing customer service on websites.
Deployment extends to mobile and web applications, making AI accessible and interactive for users.

Glossary

• Project A project is a series of tasks that need to be completed to reach a specific outcome. A project can also be defined as a set of inputs and outputs required to achieve a particular goal.
• Problem Scoping The process of finalising the aim of a system or project means you scope the problem that you wish to solve with the help of your project.
• Data visualisation it is the process of represehting data visually or graphically, by using visual elements like charts, diagrams, graphs and maps etc.
• Data Exploration The process of interpreting or exploring some useful information out of the large acquired data for a better understanding.
• Modelling The process of selecting and implementing the model which match the requirements of a project.
• Evaluation The evaluation of each and every model tried and choose the model which gives the most efficient and reliable results.
• 4W’s canvas A framework to define a problem statement using Who, What, Where\When and Why.