English | Português

Machine Learning - Dr. Diabetes - Python

Checkpoint #4 of FIAP Coding For Security

• Objective: To develop a program based on the supervised machine learning technique.
• Challenge: How do I know if I have diabetes?

---

Contents

• Group
• General Info
• Fluxogram
• Parameters
• Technologies
• Setup

---

Group

• João Pedro Zobolli Carnevalli
  • GitHub
    • @joaocarnevalli
  • Linkedin
  • Twitch
  • E-mail
    • [email protected]
• Renato Kim
  • GitHub
    • @renatokim18
  • Linkedin
• Gustavo Kondo
  • GitHub
    • @GustavoKondo
  • Linkedin
• Kaiky Amaral
  • GitHub
    • @KekDisk

---

General

This program aims to give a pre-diagnosis of diabetes based on the patient's glucose values ​​collected in three different states and situations. They are: Casual State, post overload and fasting. It can bring 3 different results.

• Normal Glucose
• Decreased Glucose
• Diabetes Mellitus

---

Fluxogram

---

Parameters

• Normal Glucose
  • Fasting:
    • less than 100mg/dL
  • Post Overload:
    • less than 140mg/dL
  • Casual blood glucose:
    • less than 200mg/dL
• Decreased Glucose
  • Fasting:
    • between 100 and 126 mg/dL
  • Post Overload:
    • between 140 to 200mg/dL
  • Casual blood glucose:
    • less than 200mg/dL
• Diabetes mellitus
  • Fasting:
    • Greater than or equal to 126mg/dL
  • Post Overload:
    • Greater than or equal to 126mg/dL
  • Casual blood glucose:
    • greater than or equal to 200mg/dL
• Criteria for evaluation
  • Fasting: No food intake for at least 8 hours
  • Post Overload: 2h after 75g of glucose
  • Casual blood glucose: Performed at any time of day

Source: GlicOnline

---

Technologies

this project was made in Python 3.10 using the scikit-learn library

---

Setup

• 1 - Install Python
• 2 - Install scikit-learn library
  • Open CMD
  • Type pip install -U scikit-learn and run it
• 3 - Download the doctormachinelearning.py and run it.