Comprehensive Study Guide for the Graduation Prediction Project Exam

Introduction to the Graduation Prediction Project Exam

• Project Context: This document serves as a comprehensive preparation guide for the SMK RPL (Software Engineering) vocational high school class promotion exam. The project centers on a graduation prediction website fueled by machine learning.

• Core Educational Focus: The material is divided into three primary pillars:

  • Machine Learning: The logic and engine of the prediction.

  • Website Development: The interface and implementation for the end-user.

  • Mathematics and Statistics: The underlying concepts and performance evaluation metrics.

Focus 1: Machine Learning Principles

• Definition of Machine Learning: Machine Learning is a specialized branch of Artificial Intelligence (AI) that allows computer systems to learn from data and formulate predictions. In the context of this graduation prediction project, the model analyzes historical student data to determine the likelihood of a student graduating or not graduating.

• The Primary Workflow: The process follows a linear progression:

  • Dataset $\rightarrow$ Training $\rightarrow$ Model $\rightarrow$ Prediction.

• Core Components and Definitions:

  • Dataset: This is the collection of historical student data.

  • Training: This is the specific process where the model undergoes learning from the provided data.

  • Model: This is the final result of the training process, which is thereafter used to perform predictions on new, unseen data.

• Dataset and Features:

  • Contents: The dataset includes variables such as daily grades ($\text{nilai harian}$), subject grades, attendance ($\text{presensi}$), and assignment completion.

  • Fitur (Features): Every column in the dataset used to generate a prediction is referred to as a feature.

  • Label or Target: The specific result that the system aims to predict is called the label or the target (e.g., "Pass" or "Fail").

• Training and Testing Data:

  • Data is typically partitioned into two main sets: data training and data testing.

  • Data Training: Used exclusively to train and teach the model.

  • Data Testing: Used to evaluate the performance and reliability of the model.

  • Objective: The goal of this split is to determine if the model is capable of predicting new data accurately and effectively.

• Accuracy and Model Evaluation:

  • Accuracy Definition: This metric indicates how many correct predictions were made in comparison to the total number of tested data points.

  • Calculation Example: If $95$ out of $100$ data points are predicted correctly, the model's accuracy is $95 \%$.

  • Caveat: While high accuracy suggests strong performance, it does not provide an absolute guarantee that every single future prediction will be correct.

Machine Learning Algorithms

• Decision Tree:

  • Functionality: This algorithm operates similarly to a decision flowchart. The model poses a sequence of questions until it reaches a final outcome.

  • Hypothetical Scenario:

  1. Is the grade above $75$?

  2. Is attendance above $80 \%$?

  3. Based on these answers, the system categorizes the student as "Pass" ($\text{Lulus}$) or "Fail" ($\text{Tidak Lulus}$).

  • Advantages: It is easy to understand, easy to visualize, and fast to execute.

  • Disadvantages: It is prone to overfitting.

• Random Forest: This consists of a collection of many Decision Trees. It is generally more stable than a single tree.

• KNN (K-Nearest Neighbors): This algorithm determines the result based on the "nearest neighbors" or the data points most similar to the new input.

• Naive Bayes: This algorithm is built upon the concepts of probability.

• Logistic Regression: This is utilized for classification and calculates the probability of a specific class occurring.

• Rationale for Multi-Algorithm Usage: Each algorithm possesses unique characteristics. By comparing several algorithms, developers can identify which model yields the highest accuracy for the specific dataset in use.

Focus 2: Website Development

• Project Architecture: The website consists of two distinct parts:

  • Frontend: The user-facing part that users interact with directly.

  • Backend: The component that processes data and communicates with the machine learning model.

• Technologies Used:

  • HTML: Used to create the fundamental structure of the web pages.

  • CSS: Used to enhance the visual aesthetics and layout.

  • JavaScript: Used to increase interactivity, such as validating user input and performing dynamic data updates.

• Backend and Flask:

  • Flask: A Python-based framework used frequently to bridge the website with the machine learning model. Flask performs the following tasks:

  1. Receives input from the user.

  2. Processes that data.

  3. Calls the machine learning model.

  4. Returns the prediction result back to the web page.

• Website Project Workflow:

  1. The user fills out a form on the frontend.

  2. The data is sent to the backend.

  3. The backend invokes the machine learning model.

  4. The model performs the prediction.

  5. The result is displayed to the user.

• Database and Data Storage:

  • Databases are utilized to store data in a structured manner.

  • In this project, student data can be stored in a database to ensure it is easily manageable and reusable for future needs.

• Comparison: Frontend vs. Backend:

  • Frontend: Focuses on the layout, visual design, and user experience ($\text{UX}$).

  • Backend: Focuses on application logic, data processing, and communication with models or databases.

Focus 3: Mathematics and Statistics

• Relevance: Machine learning is inextricably linked to statistics. Academic evaluations often connect the project to concepts such as averages, percentages, probability, and graphical analysis.

• Mean (Rata-rata):

  • Definition: The mean is obtained by summing all data points and dividing by the count of those points.

  • Formula: $\text{Mean} = \frac{\sum x}{n}$

  • Example: If grades are $80$, $90$, and $70$, the average is calculated as follows:

  • $\frac{80 + 90 + 70}{3} = 80$

• Median and Modus:

  • Median: The middle value in a dataset that has been sorted in order.

  • Modus: The value that appears most frequently in a dataset.

  • Both are fundamental concepts in basic statistics.

• Percentage ($\text{Persentase}$):

  • Frequently used to calculate accuracy.

  • Basic Formula: $\text{Percentage} = (\frac{\text{Part}}{\text{Total}}) \times 100 \%$

• Graphs and Data Interpretation:

  • Graphs are used to simplify the viewing of data comparisons, trends, and model performance.

  • On the website, graphs assist users in understanding prediction results more rapidly.

Potential Exam Questions

• What is machine learning?

• Why was the Decision Tree algorithm chosen?

• What are the definitions of dataset, training, and model?

• What is the specific function of Flask?

• What is the difference between frontend and backend?

• Why is accuracy important in this context?

• Why is it that machine learning predictions are not always correct?

• What is the relationship between this project and mathematics?

Quick Memorization Summary

• Machine Learning: Systems that learn from data.

• Dataset: Data used for training.

• Model: The result of the training process.

• Prediction: The output generated by the model.

• Frontend: Composed of HTML, CSS, and JavaScript.

• Backend: Processes data and connects the model.

• Accuracy: Indicates the level of prediction precision.

• Decision Tree: Operates like a tree of logical decisions.