ATM_cash_deposit_predictor

A platform that applies various classical machine learning algorithms and deep learning algorithms in an effort to predict the fluctuation of an ATM's cash deposit through various days of usage.

Installation

Requirements

The following software is required:

• Docker ( optionally )
• Python V3.12
• Pipenv ( if Docker is not used )
• CUDA ( optionally, to boost LSTM speed )

Configuration

After downloading the source code from git you will need to configure some parameters in the .env file. The parameters are the following:

• DATASTORE_SUPPLY_INFO : The filepath of the ATM supply info in xlsx format.
• DATASTORE_SUPPLY_TEST_INFO : The filepath of the ATM supply info for evaluation in xlsx format.
• DATASTORE_OUTAGE_INFO : The filepath of the ATM outage info.

If you do install the predictor using Docker then you will need to setup a local MongoDB and provide the following parameters:

• MONGO_USERNAME : The MongoDB username.
• MONGO_PASSWORD : The MongoDB password.
• MONGO_DATABASE : The MongoDB database.
• MONGO_AUTH_DB : The MongoDB authentication database.
• MONGO_HOST : The MongoDB host.
• MONGO_PORT : The MongoDB port.

Docker

The recomented way to install and use ATM_cash_deposit_predictor is through Docker compose. The instruction for installation are the following:

• Download the source code using git.
• Change directory to the root folder of the source.
• Execute the command docker compose up --build -d. It may need sudo rights depending on system setup.
• After the container is running you can either connect to the atmo-python container using the command docker exec -it atmo-python /bin/bash or you can run a command directly using docker exec atmo-python <command>

Pipenv

If Docker is not available or not preferred Pipenv can also be used. The instruction for installation are the following:

• Download the source code using git.
• Change directory to the root folder of the source.
• Execute the command pipenv install --python 3.12.
• After the pipenv is created you can use the pipenv shell using the command pipenv shell and execute the commands normally or execute them using pipenv run <command>.

Usage

The following commands are available:

• python main.py import_data : imports the data from the datastore files and add them to the MongoDB as defined in .env file.
• python main.py evaluate_classic_models : evaluates the classic machine learning models and prints statistics.
• python main.py evaluate_deep_models : evaluates the deep learning models and prints statistics.
• python main.py apply_models --data=”path/to/input/data” : run the prediction models on the provided dataset, defined using the --data argument, and print a list of ATMs that will need supply tommorow.

Licence

This program is free software: you can redistribute it and/or modify it under the terms of the GNU Affero General Public License as published by the Free Software Foundation, either version 3 of the License, or any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero General Public License for more details.

You should have received a copy of the GNU Affero General Public License along with this program. If not, see https://www.gnu.org/licenses/.