Cultural classifier is an application using the ML model to distinguish between two archaeological cultures (specifically epigravettian and epiaurignacian) that are hard to classify with the conventional classification approach that relies on analysing the stone tool industry.
- machine learning for classification purposes in archaeology
- practical usage of the Logistic regression model as a complement for classifying stone industry, trained model and all data are available here
- all archaeological sites are represented by using interactive maps
You can watch the video here
You can checkout the app here
pip (package manager)
python 3.6 (and newer versions)- Clone the repo
git clone https://github.com/hampet1/moodtracker.git - Install packages
pip install -r requirements.txt - Set up config
* set FLASK_ENV=development * set FLASK_APP=main - Run server
flask run
Classifier distinguishes between two very simalar archaeological cultures in terms of material culture. The input data (attributes) includes geospacial data related to the archaeological sites, the output is one of the archaeological cultures. For more information regarding input data please checkout the guideline.
For more info please checkout the website about project
The scientific paper which backs up the whole idea of applyingthe machine learning model for classification purposes in such fields as archeology is available here
It's absolutely crucial to collect more data regarding these archaeologically hard to determine cultures in order to fully understand the potential of this quite new approach. At this stage the amount of data is not sufficient, that's why we should take this project for the time being with a grain of salt.
All data was collected by associate professors Zdenka Neruda and Petr Neruda, under the patronage of the Anthropos Institute.
Distributed under the MIT License.
Petr Hamrozi - hamrozipetr@outlook.com