Forecasting Emerging Industry ETF Time Series Using ARIMA, ANN and Hybrid Models

  Рет қаралды 5,564

Andero P

Andero P

Күн бұрын

Пікірлер: 17
@AMA_RILDO
@AMA_RILDO Жыл бұрын
Very nice project, may I ask you if implementing a further LSTM, the results can improve?
@AnduOlen13
@AnduOlen13 Жыл бұрын
Substituting the MLP with a LSTM could result in better forecasting accuracy from the literature I remember reading. Adding another LSTM layer might increase the model complexity and introduce overfitting. Interesting question though and needs testing! 😄
@AMA_RILDO
@AMA_RILDO Жыл бұрын
Thank you bro for the answer 👍👍@@AnduOlen13
@bourdayrachid3855
@bourdayrachid3855 Жыл бұрын
Can you share with me the code please
@AnduOlen13
@AnduOlen13 Жыл бұрын
@@bourdayrachid3855 Github link is in the description. After the initial project, no further changes have been implemented.
@andiillaervianinensi1984
@andiillaervianinensi1984 Жыл бұрын
You have 253 row data, but when you input_dim, why only [219 rows x 1 columns]
@AnduOlen13
@AnduOlen13 Жыл бұрын
To adjust for the lags, the first 9 values (lag=9 in input_dim) were removed. 253-9=244 10% of the data was used for a test set. 244 x 0.9 = 219
@andiillaervianinensi1984
@andiillaervianinensi1984 Жыл бұрын
why you're delete of the data? Is that give the impact of your model? because we have data rainfall for prediction using hybrid ARIMA-NN but I wan't to delete my data@@AnduOlen13
@farrellubis4175
@farrellubis4175 3 жыл бұрын
I'm trying to follow your steps but I get stuck in "from Models.ARIMA import *". What module should i install to get through it ? Thanks 🙏🏻
@AnduOlen13
@AnduOlen13 3 жыл бұрын
I believe Models.ARIMA was a custom module I created that had several functions relating to ARIMA forecasting. You can find the module and its contents on my GitHub, link in the description.
@godfreyjoseph9483
@godfreyjoseph9483 3 жыл бұрын
I didn't see the forecasting part of the combined model
@AnduOlen13
@AnduOlen13 3 жыл бұрын
The forecast of the combined model plotted against the actual values can be seen at 15:00
@andiillaervianinensi1984
@andiillaervianinensi1984 Жыл бұрын
can i ask u in another sosmed?
@DataInMathematics
@DataInMathematics Жыл бұрын
Hello I try to run the code for my data but you're code is error. Can I ask u from platform instagram? Help me please
@AnduOlen13
@AnduOlen13 Жыл бұрын
Perhaps you're having problems with the different requirements. All packages are listed in the Github file "requirements.txt". There are also custom made modules in use for the project such as the contents of the "Models" folder.
@DataInMathematics
@DataInMathematics Жыл бұрын
What are the tips for setting the arima_model_order?@@AnduOlen13
My scorpion was taken away from me 😢
00:55
TyphoonFast 5
Рет қаралды 2,7 МЛН
So Cute 🥰 who is better?
00:15
dednahype
Рет қаралды 19 МЛН
Мен атып көрмегенмін ! | Qalam | 5 серия
25:41
Сестра обхитрила!
00:17
Victoria Portfolio
Рет қаралды 958 М.
7 Outside The Box Puzzles
12:16
MindYourDecisions
Рет қаралды 80 М.
MIT Introduction to Deep Learning | 6.S191
1:09:58
Alexander Amini
Рет қаралды 825 М.
Lecture on hybrid modeling and optimization of processes
34:22
Process Intelligence Research
Рет қаралды 3,4 М.
Forecasting with the FB Prophet Model
20:42
Rob Mulla
Рет қаралды 96 М.
How to build ARIMA models in Python for time series forecasting
20:38
Lianne and Justin
Рет қаралды 92 М.
What is RAG? (Retrieval Augmented Generation)
11:37
Don Woodlock
Рет қаралды 187 М.
Why Does Diffusion Work Better than Auto-Regression?
20:18
Algorithmic Simplicity
Рет қаралды 416 М.
161 - An introduction to time series forecasting - Part 1
12:58
DigitalSreeni
Рет қаралды 36 М.
Solving one of PostgreSQL's biggest weaknesses.
17:12
Dreams of Code
Рет қаралды 223 М.
My scorpion was taken away from me 😢
00:55
TyphoonFast 5
Рет қаралды 2,7 МЛН