KO
|
EN
gitlite — search
Search
#javascript
#python
#hacktoberfest
#react
#ai
#typescript
#llm
#go
#golang
#android
#machine-learning
#rust
#deep-learning
#linux
ZeroMat
★ 12
Open GitHub ↗
ZeroMat as presented at ICISCAE 2021
Download README (.md)
Explore Similar Repositories
b-u585i-iot02a-bsp
:
Provides the B-U585I-IOT02A board driver, part of the STM32Cube BSP Component for STM32U5 series.
terrapn
:
No description available.
protoc-gen-go-ddl
:
Protoc Plugin for generating DDL and Golang database models
ROME_CORE
:
No description available.
fluxt
:
Boilerplate to quickly get started on building a Flask app with a Nuxt frontend.
// repository documentation
Was this content helpful?
★ 0
(0 ratings)
Select Rating:
★
★
★
★
★
Submit Feedback
Recent Feedback
×
Download README
Do you want to download the
README.md
file for
ZeroMat
?
Download (.md)
# ZeroMat ZeroMat is an open-source software owned by Ratidar Technology LLC (http://ratidar.mysxl.cn). If you want to purchase commercial software products from us, please send an email to haow85@live.com. ZeroMat is a cold-start solver for recommender systems invented by Hao Wang and published at ICISCAE 2021. The paper can be found here : https://arxiv.org/abs/2112.03084 Please modify Line 372 for the input location of MovieLens 1M dataset's rating values if necessary. ZeroMat is really important because: 1. It is the first algorithm that solves cold start problem in recommender systems with no input data. 2. The experiments on the paper demonstrate that human cultural tastes (or, user item rating data) follows a predictable distribution after evolution for sometime. 3. Because of observations in 2, our public culture is pre-determined. Human cultural evolution converges to Zipf's distribution and the details have no relation with history. To execute ZeroMat, run the following commands : python3 process_data.py python3 compute_cold_start.py A sample plot of results would look like the following :  
