Advanced time series forecasting with PyCaret- Is stock price really predictable?

preview_player
Показать описание
Relevant playlists:
Instructor: Pedram Jahangiry

All of the slides and notebooks used in this series are available on my GitHub page, so you can follow along and experiment with the code on your own.

Lecture Outline:
0:00 is the stock price really predictable? short answer
3:35 where to find the materials
6:20 how to convert time series problem into a machine learning problem
13:00 opening the notebook in VScode (walking through the pycaret and sktime documentation)
18:40 preparing the data index (this step is important)
25:00 time series cross validation with PyCaret
43:15 Bootstrapping with PyCaret
52:30 making predictions
56:20 final words
Рекомендации по теме
Комментарии
Автор

Hello Pedram. I can't install Pycaret on VSC. I'm trying to replicate your code on VSC using Jupyter notebook. However, Pycaret can't be installed. On the other hand, how can I change your data base used in Google Colab? I want to use Data from TSLA or NVDA.

aarondelarosa
Автор

Hello Pedram, I wanted to know if I can assign weights to my data such that there's more weightage to the recent data and less to the older data. Could you please help me this. Thank you!

dimpleraghu
Автор

I want to ask:

1) Is it necessary to have stationary time series or not?
2) If we make our data series data frame of "df" that included the stock prices of AAPL and of TSLA, and if we put target='AAPL' in our setup, will the program consider the values of TSLA as an explanatory time series or not?

Thank you very much!

pepe_the_frog-
Автор

Kind of unrelated question but what do you use to have that Jupyter notebook interactable graphs, looks amazing. Thank you for the video, a lot of valuable information!

metinunlu_
Автор

Hi Pedram, I really need your help.
I'm working on a project that detects seasonality using pycaret setup and it's all working great except for that the setup function shows seasonality present even with the most randomest data eg:- just a data of random values ranging from 20-25. Is there a way to fix this?

dimpleraghu
Автор

Great vídeo Pedram!

Will you make more videos about time series with PyCaret?

I want to predict monthly energy market consumption, but I need several explanatory time series to have e decent result.

In PyCaret, do the explanatory variables require future values? For example, GDP is a crucial explanatory variable, but it is challenging to forecast. Must I have a GDP projection to use it as an explanatory variable to predict monthly energy market consumptuon?

vitorbarros
Автор

Very nice lecture. How did you configure VS code to show pycaret objects output? Mine does not show (I had to use Jupyter lab/Colab to see the table outputs). Can you do a video about blending? Taking top 3 models and blending them... also, why did you decide to use sliding?

Darokbr
Автор

Excellent. Could you help me out to complete a Python project using NeuralProphet.? It seems that NeuralProphet can't be correctly installed. There's something wrong with the NeuralProphet model.

aarondelarosa
Автор

Hello,
I have a question about PyCaret. When employing `create_model`, is the model generated with random hyperparameter values, or is there a specific method for their selection?

AliGolafshan-wi
Автор

Pycaret is awesome for model selection. But please make the code area bigger. Cant see.

superfreiheit
Автор

Omg, why so sucky, 1 hour of excuses no u cant

MarkWeits