Upload 7 files
Browse files- README.md +150 -1
- feature_columns.pkl +3 -0
- kmeans.pkl +3 -0
- nlp_model.pkl +3 -0
- risk_model.pkl +3 -0
- spending_model.pkl +3 -0
- tfidf.pkl +3 -0
README.md
CHANGED
|
@@ -1,3 +1,152 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Expense AI Intelligence System
|
| 2 |
+
|
| 3 |
+
**Author:** Krishnamohan Yagneswaran
|
| 4 |
+
|
| 5 |
---
|
| 6 |
+
|
| 7 |
+
## Overview
|
| 8 |
+
|
| 9 |
+
This project is a machine learning system that analyzes expense data and provides useful financial insights. It combines multiple models to predict spending, detect risk, classify transactions, and understand user behavior.
|
| 10 |
+
|
| 11 |
+
The system is designed to work with real-world expense data and help users make better financial decisions.
|
| 12 |
+
|
| 13 |
---
|
| 14 |
+
|
| 15 |
+
## What This Model Does
|
| 16 |
+
|
| 17 |
+
This system provides the following outputs:
|
| 18 |
+
|
| 19 |
+
* Predicts how much money you may spend
|
| 20 |
+
* Identifies if a transaction is high risk or low risk
|
| 21 |
+
* Classifies the type of expense (food, travel, bills, etc.)
|
| 22 |
+
* Groups spending behavior into clusters
|
| 23 |
+
* Provides a simple financial decision suggestion
|
| 24 |
+
|
| 25 |
+
---
|
| 26 |
+
|
| 27 |
+
## About the `.pkl` Files
|
| 28 |
+
|
| 29 |
+
All the `.pkl` files in this repository are **trained machine learning models and components**. These files are saved using `joblib` and allow the system to be reused without retraining.
|
| 30 |
+
|
| 31 |
+
### Files Explanation:
|
| 32 |
+
|
| 33 |
+
* `spending_model.pkl`
|
| 34 |
+
→ Predicts future spending amount
|
| 35 |
+
|
| 36 |
+
* `risk_model.pkl`
|
| 37 |
+
→ Classifies whether spending is HIGH or LOW risk
|
| 38 |
+
|
| 39 |
+
* `nlp_model.pkl`
|
| 40 |
+
→ Predicts category from text input
|
| 41 |
+
|
| 42 |
+
* `tfidf.pkl`
|
| 43 |
+
→ Converts text into numerical format for NLP
|
| 44 |
+
|
| 45 |
+
* `kmeans.pkl`
|
| 46 |
+
→ Groups spending behavior into clusters
|
| 47 |
+
|
| 48 |
+
* `feature_columns.pkl`
|
| 49 |
+
→ Stores the correct input structure used during training
|
| 50 |
+
|
| 51 |
+
---
|
| 52 |
+
|
| 53 |
+
## How to Use
|
| 54 |
+
|
| 55 |
+
### 1. Install Required Libraries
|
| 56 |
+
|
| 57 |
+
```bash
|
| 58 |
+
pip install pandas numpy scikit-learn joblib
|
| 59 |
+
```
|
| 60 |
+
|
| 61 |
+
### 2. Load Models
|
| 62 |
+
|
| 63 |
+
```python
|
| 64 |
+
import joblib
|
| 65 |
+
|
| 66 |
+
spending_model = joblib.load("spending_model.pkl")
|
| 67 |
+
risk_model = joblib.load("risk_model.pkl")
|
| 68 |
+
nlp_model = joblib.load("nlp_model.pkl")
|
| 69 |
+
vectorizer = joblib.load("tfidf.pkl")
|
| 70 |
+
kmeans = joblib.load("kmeans.pkl")
|
| 71 |
+
```
|
| 72 |
+
|
| 73 |
+
### 3. Example Input
|
| 74 |
+
|
| 75 |
+
```python
|
| 76 |
+
input_data = {
|
| 77 |
+
"weekday": 2,
|
| 78 |
+
"rolling_avg": 500
|
| 79 |
+
}
|
| 80 |
+
```
|
| 81 |
+
|
| 82 |
+
### 4. Predict
|
| 83 |
+
|
| 84 |
+
```python
|
| 85 |
+
prediction = spending_model.predict([list(input_data.values())])
|
| 86 |
+
print(prediction)
|
| 87 |
+
```
|
| 88 |
+
|
| 89 |
+
---
|
| 90 |
+
|
| 91 |
+
## System Features
|
| 92 |
+
|
| 93 |
+
* Uses multiple machine learning models together
|
| 94 |
+
* Works with structured CSV expense data
|
| 95 |
+
* Includes NLP for text-based classification
|
| 96 |
+
* Uses clustering for behavior analysis
|
| 97 |
+
* Provides decision-making suggestions
|
| 98 |
+
|
| 99 |
+
---
|
| 100 |
+
|
| 101 |
+
## Use Cases
|
| 102 |
+
|
| 103 |
+
* Personal finance tracking
|
| 104 |
+
* Budget planning
|
| 105 |
+
* Expense monitoring applications
|
| 106 |
+
* Financial analytics systems
|
| 107 |
+
* Academic and learning projects
|
| 108 |
+
|
| 109 |
+
---
|
| 110 |
+
|
| 111 |
+
## License
|
| 112 |
+
|
| 113 |
+
This project is licensed under the **MIT License**.
|
| 114 |
+
|
| 115 |
+
You are free to:
|
| 116 |
+
|
| 117 |
+
* Use
|
| 118 |
+
* Modify
|
| 119 |
+
* Distribute
|
| 120 |
+
* Use commercially
|
| 121 |
+
|
| 122 |
+
As long as the original license and author credit are included.
|
| 123 |
+
|
| 124 |
+
---
|
| 125 |
+
|
| 126 |
+
## Credits
|
| 127 |
+
|
| 128 |
+
If you use this project in your work, research, application, or product, it would be appreciated if you provide credit:
|
| 129 |
+
|
| 130 |
+
**Krishnamohan Yagneswaran**
|
| 131 |
+
|
| 132 |
+
This helps support further development and recognition of the project.
|
| 133 |
+
|
| 134 |
+
---
|
| 135 |
+
|
| 136 |
+
## Future Improvements
|
| 137 |
+
|
| 138 |
+
* Web application interface
|
| 139 |
+
* Mobile application integration
|
| 140 |
+
* API deployment
|
| 141 |
+
* Real-time analytics
|
| 142 |
+
|
| 143 |
+
---
|
| 144 |
+
|
| 145 |
+
## Conclusion
|
| 146 |
+
|
| 147 |
+
This project demonstrates how multiple machine learning techniques can be combined into a single intelligent system. It is simple, practical, and scalable for real-world applications.
|
| 148 |
+
|
| 149 |
+
---
|
| 150 |
+
|
| 151 |
+
**Created by:**
|
| 152 |
+
Krishnamohan Yagneswaran
|
feature_columns.pkl
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a9651251f2e38d0eba38a3450541c580915e34709a8bada222eac19ab8865e27
|
| 3 |
+
size 168
|
kmeans.pkl
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5825f8a0555c3c3679ab2096d227096e1d9153668d1e0799b5c3faedf9d50e35
|
| 3 |
+
size 5847
|
nlp_model.pkl
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4e460bf95c74634e876492c5f75774bcda995210248ced5ce2b4e2add4d44145
|
| 3 |
+
size 48640
|
risk_model.pkl
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:685228c643bcd3b04c0bf6dacc359d7581f977aca6e015b3ee47e7861cb81ca0
|
| 3 |
+
size 56763
|
spending_model.pkl
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:69c5e7d3e07d3c72dd9f39339d292eededcfa58e6e3e90808e1c712e98b82893
|
| 3 |
+
size 142568
|
tfidf.pkl
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:126f32bd15c699394ce2a089025001352276227ac08888a6575ca6e29dbed22a
|
| 3 |
+
size 1055
|