davide-chiuchiu Goto Github PK

analysis_of_job_opening_rejections

With this project I analyzed the emails where I got job rejections with python and NLP. From this analysis I found out that emails where companies reject a candidacy and the emails where companies acknowledge a submitted candidacy have the same length, but use different keywords. I also found out that larger companies tend to reject candidacies faster than small-to-medium enterprises (10 days and 20 days respectively).

basic_tensorflow_mwe_for_animal_adoption

This repository contains the code to produce a simpletensorflow classification model with hyperparameter auto-tuning. The code in here contains a bare-bone minimum on all the fundamental steps needed to build a tensorflow model which predicts which animals will be adopted based on various features. As such, it's a reasonable starting point for more complex projects.

bigquery_on_stackoverflow_data

This project fetches data from the public stackoverflow database hosted in google cloud with bigquery and creates a dashboard with two insights: (1) the cumulative number of questions for the 8 topics with the most common tags in the database, and (2) the probability to get the accepted answer to a question within hours, days, weeks or months after it has been posted for the the 8 topics with the most common tags in the database.

chicago_crimes

Building a predictive model of arrests based on the chicago crime dataset with tensorflow

companies_often_give_the_wrong_name_to_job_openings

This project shows that companies often give the wrong name to different IT jobs. As a consequence, companies may fail to attract good candidates because an applicant has a significant probability to apply for the wrong job.

data_cleaning_korea_cases_2

data_cleaning_tv_commercials

deployment_with_ansible_and_docker

This project contains the essential files to deploy a docker container on a server with ansible. The docker container hosts an html page via an nginx server.

lex-arcana-statistic

A repo with code for analysis of success probability in the RPG Lex Arcana

music_recommendation_engine_with_tensorflow

This project summarizes the basic steps required to implement a basic recommendation engines that suggests new bands to users. Data are fetched from the open dataset of ListenBrainz in Bigquery. The recommendation engine is built by hacking the keras embedding layers to perform matrix factorization.

python_sample_with_forest_fires

This project is a showcas of the data science workflow in python. The dataset is a collection of fire sizes in forests based on several indicators. The analysis failed to show correlations between fire indicators and fire size, but it gives a basic workflow in python.

sier_covid