Upload 3 files

README.md

@@ -1,3 +1,49 @@
----
-license: apache-2.0
----
+Fake Profile Detection 
+# 🚨 Fake Instagram Profile Detection using Machine Learning
+
+This project is a real-time Instagram profile analyzer that predicts whether a given profile is **fake** or **real** using machine learning. It uses profile metrics like follower count, following count, post count, and verification status to make predictions.
+
+---
+
+## 📌 How It Works
+
+- You enter an Instagram **username**.
+- The application uses the **Apify API** to fetch public profile data.
+- It extracts key features such as:
+  - Number of followers
+  - Number of followings
+  - Number of posts
+  - Is the account private?
+  - Is the account verified?
+- These features are passed into a pre-trained **machine learning model** (`classifier.pkl`) to predict whether the profile is real or fake.
+
+---
+
+## 🛠 Technologies Used
+
+- **Python**
+- **Streamlit** – for building the web app
+- **Joblib** – for loading the ML model
+- **Apify API** – to scrape Instagram data
+- **Scikit-learn** – for training the ML model
+- **Pandas, NumPy** – for data manipulation
+
+---
+
+## 🧠 ML Model
+
+The model is trained using a labeled dataset containing Instagram profile attributes. The classification is binary:
+- `0` → Likely Fake
+- `1` → Likely Real
+
+The training includes feature normalization and multiple algorithm trials like Logistic Regression, Decision Trees, and Random Forests. The final deployed model is chosen based on accuracy and generalization.
+
+---
+
+## 🖥️ Project UI
+
+- The app is built with **Streamlit** for a clean and interactive interface.
+- Users simply input a **username** and click **Predict**.
+- Output shows the profile’s stats and the prediction result with appropriate messaging (Success/Error).
+
+---

classifier.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b2ad437a3a0a74a8f793b60926ac67c57912c9a4bc08d9cb480e99cc8662d09
+size 413195

fpd1.py

@@ -0,0 +1,39 @@
+import numpy as np
+import pandas as pd
+import streamlit as st 
+import joblib
+from apify_client import ApifyClient
+model = joblib.load("classifier.pkl")
+client = ApifyClient("apify_api_nscRkHOyMh3mytIWftXpHpZlIzBhgF4mZyPV")
+st.title("Fake Instagram Profile Detection")
+st.write("Plaese provide instagram account details you would like to predict")
+n = st.text_input("Enter username ")
+run_input = { "usernames": [n] }
+run = client.actor("dSCLg0C3YEZ83HzYX").call(run_input=run_input)
+m = client.dataset(run["defaultDatasetId"])
+for item in m.iterate_items():
+        postsCount= item.get('postsCount')
+        followersCount = item.get('followersCount')
+        followsCount = item.get('followsCount')
+        private=item.get('private')
+        verified=item.get('verified')
+
+def predictor(postsCount,followersCount,followsCount,private,verified):
+    prediction = model.predict([[postsCount,followersCount,followsCount,private,verified]])
+    print(prediction)
+    return prediction
+
+
+if st.button("Predict"):
+    result = predictor(postsCount,followersCount,followsCount,private,verified)
+    st.write("The number of posts : " , postsCount)
+    st.write("The number of followers : " ,followersCount)
+    st.write("The number of following : " ,followsCount)
+    st.write("Private : " ,private)
+    st.write("Verified : " ,verified)
+    if postsCount == None:
+        st.error("The User Doesn't exist")
+    elif result == 0 and postsCount != None:
+        st.error("The Account is Likely to be Fake ")
+    else:
+        st.success("The Account is Likely to be Real")