Penny Investment Manager

Training, Validation, and Implementation

This repository provides the backbone for the Penny Investment Management code. Included is the scripts to train, validate and test edge cases, and then implement into usage.

---

Technical Overview & Methodology

The core architecture operates across three interconnected modules: predictive feature construction, historical portfolio simulation, and live signal dispatch.

+-----------------------------------------------------------------------------------+
|                           1. MODEL PREDICTIVE PIPELINE                            |
|                                                                                   |
|  Historical Market Data ---> Feature Vector Construction ---> Probabilistic Score |
+-----------------------------------------------------------------------------------+

|
v

+-----------------------------------------------------------------------------------+
|                        2. SIMULATION & BACKTESTING ENGINE                         |
|                                                                                   |
|  Signal Evaluation ---> Threshold Scaling ---> Ledger & Capital Reallocation      |
+-----------------------------------------------------------------------------------+

|
v

+-----------------------------------------------------------------------------------+
|                      3. REAL-TIME INFERENCE & ALERT SERVICE                       |
|                                                                                   |
|  Live Data Ingestion ---> Dynamic Feature Tensor ---> Alert & Diagnostic Delivery |
+-----------------------------------------------------------------------------------+

See full overview in this documentation.

System Usage & Execution Workflow

Deploying Docker Container

Grab dockerfile and initialize container

docker build -t penny_build https://raw.githubusercontent.com/mastermind-mayhem/penny/deploy/Dockerfile
docker run --name penny -d penny_build

Configure config.ini to customize what stocks to watch and what confidence level to have:

docker exec -it penny bash
sudo apt-get update
sudo apt-get install -y nano
sudo nano /opt/penny/config.ini

Installing Individually on Linux

Pull the install script from the repository

curl -O https://raw.githubusercontent.com/mastermind-mayhem/penny/deploy/install.sh

Recognize as a Shell Script and execute

sudo chmod +x install.sh
sudo ./install.sh

Configure config.ini to customize what stocks to watch and what confidence level to have:

sudo nano /opt/penny/config.ini

Technical Stack

• Language: Python 3.9+
• Deep Learning Framework: TensorFlow / Keras
• Numerical Processing & Feature Extraction: NumPy, Pandas, Scikit-Learn
• Visualization Engine: Matplotlib
• Market Data Ingestion: Open-source financial APIs
• Alert Delivery: ntfy HTTP protocol (compatible with iOS and Android)

---

License

Distributed under the MIT License.