Using shallow (as opposed to deep) models, develop predictive models that predict the temperature (Temperatura) on the next seventh day:
* Don't use forecasting models like ARIMA or Holt-Winters
* Set variables from historical data. For example, you can set data such as the average temperature of the last 3 days, of the last 7 days, of the last 10 days, the minimum air pressure of the last 3 days, of the last 7 days, of the last 15 days , the values of the previous day of all the variables, the values of 1 year ago on the same date, etc.
* You must create a cross section dataset, estimating the variables to use each day, or each week, or each month.
* Use models like SVM and traditional artificial neural networks
* Use variable selection techniques such as PCA, LDA and Lasso
* Try applying different normalization techniques
24 pekerja bebas membida secara purata $12/jam untuk pekerjaan ini
Hello, I'm senior Data scientist i read the projects details and data attached and i'm ready to start, Also i did the same project on Kaggle before contact me and let's discuss through chat Best regards