time-series-4 models to try for every problem

1. holtwinters

1.1. https://medium.com/analytics-vidhya/python-code-on-holt-winters-forecasting-3843808a9873 https://www.kaggle.com/code/prakharprasad/smoothing-holt-winters-forecast https://www.statsmodels.org/dev/generated/statsmodels.tsa.holtwinters.ExponentialSmoothing.html https://www.analyticsvidhya.com/blog/2021/08/holt-winters-method-for-time-series-analysis/ https://timeseriesreasoning.com/contents/holt-winters-exponential-smoothing/ https://www.bounteous.com/insights/2020/09/15/forecasting-time-series-model-using-python-part-two

2. Sarimax

2.1. https://analyticsindiamag.com/complete-guide-to-sarimax-in-python-for-time-series-modeling/ https://www.statsmodels.org/dev/generated/statsmodels.tsa.statespace.sarimax.SARIMAX.html https://towardsdatascience.com/time-series-forecasting-with-arima-sarima-and-sarimax-ee61099e78f6 https://www.jadsmkbdatalab.nl/forecasting-with-sarimax-models/ https://geometrein.medium.com/time-series-forecasting-with-sarimax-c352e4f2012e PMDARIMA ----> from pmdarima.arima import auto_arima https://medium.datadriveninvestor.com/time-series-prediction-using-sarimax-a6604f258c56 http://alkaline-ml.com/pmdarima/ https://snyk.io/advisor/python/pmdarima/functions/pmdarima.auto_arima https://stevenkyle2013.medium.com/quick-intro-auto-arima-from-pmdarima-package-e7aab5e8dfb8

3. Four different models to try everytime for time series problems

4. Example of a prophet model for monthly forecast form chatGPT from fbprophet import Prophet import pandas as pd # Load the data into a pandas DataFrame df = pd.read_csv("monthly_data.csv") # Rename the columns to match the format required by Prophet df = df.rename(columns={"date": "ds", "value": "y"}) # Create a Prophet model m = Prophet() # Fit the model to the data m.fit(df) # Create a future dataframe with a range of dates future = m.make_future_dataframe(periods=12, freq='M') # Make predictions forecast = m.predict(future) # Plot the forecast m.plot(forecast)

5. LSTM

5.1. http://colah.github.io/posts/2015-08-Understanding-LSTMs/ LSTM using scalecast https://towardsdatascience.com/exploring-the-lstm-neural-network-model-for-time-series-8b7685aa8cf These are all from a same guy for medium Looks like these use the test infromation to predict test data - so there is data leakage https://machinelearningmastery.com/time-series-prediction-lstm-recurrent-neural-networks-python-keras/ https://machinelearningmastery.com/how-to-develop-lstm-models-for-time-series-forecasting/ https://machinelearningmastery.com/how-to-develop-lstm-models-for-multi-step-time-series-forecasting-of-household-power-consumption/ https://machinelearningmastery.com/multivariate-time-series-forecasting-lstms-keras/ More LSTM https://medium.com/@cdabakoglu/time-series-forecasting-arima-lstm-prophet-with-python-e73a750a9887 https://www.kaggle.com/code/ritesh7355/develop-lstm-models-for-time-series-forecasting https://blog.eduonix.com/system-programming/time-series-prediction-using-lstm/ https://www.tutorialspoint.com/time_series/time_series_lstm_model.htm https://www.tensorflow.org/guide/keras/rnn

6. Scalecast forecasting library seems pretty awesome for quick forecasting checks https://pypi.org/project/SCALECAST/ https://towardsdatascience.com/introducing-scalecast-a-forecasting-library-pt-1-33b556d9b019 https://towardsdatascience.com/may-the-forecasts-be-with-you-introducing-scalecast-pt-2-692f3f7f0be5 https://towardsdatascience.com/forecast-on-3-levels-introducing-scalecast-pt-3-eb725e0be6c9

7. Prophet

7.1. https://facebook.github.io/prophet/docs/quick_start.html#python-api https://machinelearningmastery.com/time-series-forecasting-with-prophet-in-python/ https://www.kaggle.com/code/prashant111/tutorial-time-series-forecasting-with-prophet Comparison with Arima and LSTM https://neptune.ai/blog/arima-vs-prophet-vs-lstm