Project Abstract: Design Thinking & Problem Solving

• Project Abstract Creation utilizes the Design Thinking Framework for problem-solving and detailing.

Format for Project Abstract Creation

• Project Name, Team Members Name, Problem Selection (aligned with SDGs and solvable with AI), Users Affected, Empathise (Empathy Map), Define the Problem (5W1H questions), Ideate (Brainstorm), Prototype (simple drawing, poster, gadget), and Test (optional).

Example Project Abstract

• Project Name: Suitable name related to problem and creativity.
• Team Members Name: Students can fill themselves.
• Problem Selection: Class X students often confused about which subjects to choose in the future based on a survey.
• Issue is taking admission to plus-two courses are often confused with subject selection, the issue is aligned to SDG-4, Quality Education

Users

• Class X students face challenges selecting appropriate subjects for higher studies.

Empathise

• Interviewing plus-two students reveal difficulties in subject selection and school.
• Interviewing prospective students reveals anxieties and expectations.

Define: 5W1H Questions & Answers

• Who: Students seeking plus-two admission and their parents or well-wishers.
• What: The Problem of not being able to select subject of choice, and the choice of school.
• Where: Problem occurs in all admission places.
• When: The problem occurs during admission time.
• Why: The problem occurs because of lack of knowledge of available subjects. Solving problem would result in society only with aptitudes build in the future.
• How: A system could analyze student aptitude and suggest subjects and compatible schools, giving effective education..

Problem Statement

• Students struggle choosing subjects for plus-two courses, it can be solved by a system analyzing aptitude and suggesting subjects.

Ideate

• Includes creating an application for subject suggestions based on interest.
• A chatbot for conversation, a robot for discussion and advice.
• A mobile application aggregates queries from different sources

Prototype

• Creating a chatbot to offer solutions to students for higher education.

Data Literacy

• Data is defined as facts or instructions about an entity that can be processed.
• AI is data-driven where raw data is converted into actionable, usable information.
• Data literacy includes collecting, organizing, analyzing data, checking quality, and using it ethically.
• AI data analysis uses techniques and data science to improve cleaning, inspecting, and modeling both structured and unstructured data.
• Data helps to make decisions and drawing conclusions.

Data Collection

• Allows capturing a record of past events to find recurring patterns
• From patterns can build predictive models using machine learning algorithms that look for trends and predict future changes.
• Involves pooling data by capturing it from multiple online and offline sources.
• Data volume depends on dataset features as more data leads to better predictions.
• Data diversity is most important to cover more scenarios under which model wil be used.
• Data quantity also depends on the complexity of a model from small to large data.
• A data scientists should first understand the problem and prefer solution and the data requirements
• Data is the core part of any project so it is identified, collected, and analyzed iteratively.
• Data collection happens through Primary and Secondary sources.

Primary Sources

• Sources where data is created to collect data for analysis using following methods

Survey

• Useful for measuring opinions and demographics through interviews, questionnaires, or online forms.

Interview

• Gathers information through direct communication with individuals or groups and can be structured.

Observation

• Records behaviors or events, often used in ethnographic research where direct interaction isnt possible

Experiment

• Manipulates outcomes to establish cause-and-effect relationships.

Marketing Campaign

• Uses customer data to predict behavior and optimize campaign performance.

Questionnaire

• Collects quantitative and qualitative data through structured questions, with options for open-ended feedback.

Secondary data

• Data already available, such as in books, journals, databases, to reuse for analysis

Social media data tracking

• Analyzes user generated content to understand use reception towards a new product launch.

Web scraping

• Automated tools extract content and data from websites to compare product prices.

Satellite data tracking

• Gathers earth and atmosphere data for monitoring environmental changes.

Online data platforms

• Websites offer pre compiled data like Kaggle and Github.

Exploring Data

• Data is about "getting to know" the data and its values if typical or not and correct any problem that would affect drawn conclusions.

Levels of Measurement

• It is classifying datasets based on if are quantitative or qualitative, and whether continuous or discrete datasets
• Qualitative can be nominal or ordinal, and quantitative can be interval and ratio.

Nominal level

• Nominal variables are categories which dont rank like car models or seasons.

Ordinal Level

• Ordinal data are categories that follow in order or rank, but their differences aren't calculated.

Interval level

• Similar to ordinal data in ordering, can measure differences, but no zero value.

Ratio scale level

• Is like interval data but with zero point so ratios can be calculated for better comparisons.

Central Tendency

• Statistics summarize a dataset with a single value representing entire data with single value
• Can be done with Python libraries

Mean

• Average estimated obtained dividing total of the values of a variable by number of observations

Median

• Divides the data value by equal parts where comprises greater and smaller median values

Mode

• Represented by highest bar chart that occurs most frequently in the data series

Standard Deviation

• Variance and Measures the Dispersion and provides info of the spread of data around the center point.
• Small variance means data points close to the mean
• High variance mean data points are spread out far from the mean
• Low means the data points tend to be VERY close to the mean
• High mean the data points spread out over a large range of values

Data representation

• Statistics that's a discipline concerns interpretation and presentation of the data by statistics.

Non Graphical Technique

• It can be done with the tabular or case form which is rarely used since its not useable for large datasets when analysing them
• Not effective for decisions

Graphs

• The visual display is easier for the human brain to read using lines dots etc
• Data Visualization:
• Line graphs
• Bar diagrams
• Pie diagram
• Scatter Plots
• Histogram

Line Graph

• Powerful tool displaying continuous data along a numbers axis.

Bar Graph

• Graph presenting categorical data for comparison using rectangular bars

Histogram

• Graph depicts data with various value ranges and can only represent one data axis