ML.NET Forcasting Light Switches on and off using TimeSerie or Regression once a week [Question] · dotnet/machinelearning-samples#846

(1 评论) (0 反应) (0 负责人)PowerShell (4,683 star) (2,681 fork)batch import

P3help wantedquestion

描述

Question: i have data that has light IDs and when a light of a particular ID is switched on the data is captured now the next step is to try and figure out which light will be on for a particular week in the future. Can anyone let me know how this can be done either using Time series or regression link to the data

all i need to predict is an On or Off like a for the next week or 2

Key Light ID: this is a light represented ad an ID LightDate: this is data captured once a week over the course of 2 n half years OnOff: If the light for a particular ID was on on that time or off On being 1 and off Being 0 WeightByMonth: how many times the light was on for that month max can only be 4 since the data is captured once a week WeightByYear: how many times the light was on for that year TotalWeightOverTheYears: how many times the light was on for the entirety of the data being captured

贡献者指南

技术栈: csharp
领域: machine learningdata
议题类型: research
难度: 4
预计时间: over 1 week
活动状态: stale
清晰度: unclear
前置要求: basic machine learning conceptsC# programming
新手友好度: 20
研究方向: This issue is a question about forecasting light switches using ML.NET time series or regression. To contribute, first clarify the acceptance criteria with maintainers, as the current description lacks a specific request. Then, explore ML.NET's time series methods (e.g., Singular Spectrum Analysis) and suitable regression approaches. Review existing samples at dotnet/machinelearning samples for similar patterns. Implement a solution that predicts On/Off for a given week using the provided dataset. The deliverable could be a sample notebook or code snippet.