Time series forecasting machine learning model
Budget: $250 – $750 USD
I am looking for a skilled data scientist to develop a time series forecasting machine learning model. The project involves forecasting for the next 2 periods with an achievement of MAE of less than 50.
Skills and Experience:
- Strong knowledge and experience in developing machine learning models for time series forecasting
- Proficiency in handling and analyzing historical sales data
- Ability to understand and interpret complex patterns and trends in data
- Expertise in selecting and implementing appropriate machine learning algorithms
- Familiarity with techniques for optimizing and fine-tuning models
- I use Python to code, i would appreciate it if the model is coded in python
I am not really interested in the candidate's track record, i just want a reliable forecasting model that can predict the 2 periods with an MAE of less than 50. Its up to you on how you split your percentage training and testing data, whether its 80%-20%, 90%-10%, 95%-5%, etc.
Data Description.
- There are 550 data points.
- There pattern is not see to easy, hence it requires a good machine learning model/any model applicable to learn the hidden pattern.
- The 550 data points are the historical/previous moves made by an chess game, the chess can make a total of 376 992 non-repetitive moves. Each move is assigned a value between 1 and 376 992, each data point represent the assigned value of the move made.
- I have attached the image of the 550 data points as well the excel file with the 550 data points.
Good luck!!! :)
Skills and Experience:
- Strong knowledge and experience in developing machine learning models for time series forecasting
- Proficiency in handling and analyzing historical sales data
- Ability to understand and interpret complex patterns and trends in data
- Expertise in selecting and implementing appropriate machine learning algorithms
- Familiarity with techniques for optimizing and fine-tuning models
- I use Python to code, i would appreciate it if the model is coded in python
I am not really interested in the candidate's track record, i just want a reliable forecasting model that can predict the 2 periods with an MAE of less than 50. Its up to you on how you split your percentage training and testing data, whether its 80%-20%, 90%-10%, 95%-5%, etc.
Data Description.
- There are 550 data points.
- There pattern is not see to easy, hence it requires a good machine learning model/any model applicable to learn the hidden pattern.
- The 550 data points are the historical/previous moves made by an chess game, the chess can make a total of 376 992 non-repetitive moves. Each move is assigned a value between 1 and 376 992, each data point represent the assigned value of the move made.
- I have attached the image of the 550 data points as well the excel file with the 550 data points.
Good luck!!! :)
Related categories:
Machine Learning (ML)
Deep Learning
Data Modeling
Data Engineer
Time Series Forecasting