This is a machine learning project of being able to predict whether a transaction is a fraud or not. Take note the Dataset is taken from zindi from a closed competition.
1. Python 3.7
2. pip
3. Virtualenv(python package)
- Create Virtual environment.
on Unix based System
python3 -m venv envor on Win32
py -m venv env
- Switch to Virtual environment
on Unix based System
source env/bin/activate
or Win32
.\env\Scripts\activate
- Change KERAS_BACKEND to either tensorflow if you your CPU supports AVX otherwise Shift to theano for leagcy CPU's
on Unix Based System
export KERAS_BACKEND=theano
on Windows Based System
set KERAS_BACKEND=theano
- Create a Model
python3 train.py
- Infer Model
python3 main.py infer json [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13]
Where the Arguments are as follows:
[0] = TransactionId [1] = BatchId [2] = AccountId [3] = SubscriptionId [4] = CustomerId [5] = CurrencyCode [6] = ProviderId [7] = ProductId [8] = ProductCategoryId [9] = ChannelId [10] = Amount [11] = Value [12] = TransactionStartTime [13] = PricingStrategy
Enjoy