Thank you 😍I have 2 questions. Can we do this on an hourly basis? Also, Have you looked at the correlation between variables in your previous videos? I want to watch
@@onurdatascience My data is hourly. I have seasonality both during the day and during the year. According to my research, there may be a noise problem in the model, right?
Yes, when dealing with hourly data that exhibits both daily and yearly seasonality, it's common to encounter noise in the model due to the complexity and variability of these patterns. This noise can arise from factors such as unaccounted external influences and fluctuations within and across days and years. To address this issue, it's essential to select appropriate modeling techniques like SARIMA that can capture the intricate seasonal patterns effectively. Additionally, feature engineering, regularization methods like ridge regression, and ensemble approaches such as random forest can help mitigate noise and improve the overall performance of the model by extracting relevant features, penalizing complexity, and combining multiple models.
@@onurdatascience Sarimax may be the solution, but I know that it can only have 1 external variable. Can I make more than 1 external variable? Thank you for your answers
thank you, I appreciate if you can make a video to use LSTM for transfer leaning for unseen data
Thank you, of course I can. I added it to my future videos list
Thank you 😍I have 2 questions. Can we do this on an hourly basis? Also, Have you looked at the correlation between variables in your previous videos? I want to watch
Hello. Yes we can do this on hourly basis. Yes I generally check for correlation in data analysis project videos. Thanks for watching!
@@onurdatascience My data is hourly. I have seasonality both during the day and during the year. According to my research, there may be a noise problem in the model, right?
Yes, when dealing with hourly data that exhibits both daily and yearly seasonality, it's common to encounter noise in the model due to the complexity and variability of these patterns. This noise can arise from factors such as unaccounted external influences and fluctuations within and across days and years. To address this issue, it's essential to select appropriate modeling techniques like SARIMA that can capture the intricate seasonal patterns effectively. Additionally, feature engineering, regularization methods like ridge regression, and ensemble approaches such as random forest can help mitigate noise and improve the overall performance of the model by extracting relevant features, penalizing complexity, and combining multiple models.
@@onurdatascience Sarimax may be the solution, but I know that it can only have 1 external variable. Can I make more than 1 external variable? Thank you for your answers