Optimizing architectures of recurrent neural networks for improving the accuracy of time series forecasts

Июль 21, 2022

Главная
Английский язык
Optimizing architectures of recurrent neural networks for improving the accuracy of time series forecasts

Содержание

2. Outline The main purpose and subtask of research Literature review results Comparison of different models on
3. The purpose and subtasks The main purpose: Optimizing RNN parameters to improve the accuracy of forecasting
4. Literature review Common approaches for analysing financial time series: 1) Classic statistical methods Regression models Autoregressive
5. Literature review Specific features of statistical approaches: Demonstrate high accuracy result especially when time series have
6. Literature review Specific features of Recurrent Neural Networks: Able to approximate complex relationships in time series
7. Literature review Long Short-Term Memory extends the RNN architecture with a standalone memory Fig. 1 –
8. Literature review Feature based approach to model selection 8/17
9. Literature review Algorithm of model selection 9/17
10. Model comparison Training details: Linear Regression - 91 parameters including bias Holt-Winters Model - alpha =
11. Model comparison Data description Fig. 1 - Daily transaction amount, millions of rub (from Nov. 2013
12. Model comparison Results Fig. 2 – Linear regression 12/17
13. Model comparison Results Fig. 3 – Holt-Winters Model 13/17
14. Model comparison Results Fig. 4 – Recurrent Neural Network 14/17
15. Model comparison Results Table 1 – Performance of models on test data set 15/17
16. Outputs Findings: Recurrent Neural Network can outperform the other classical statistical models in predictive accuracy More
18. Скачать презентацию

Слайд 2

Outline
The main purpose and subtask of research
Literature review results
Comparison of

different models on transaction data
Further work plan

2/17

Слайд 3

The purpose and subtasks
The main purpose:
Optimizing RNN parameters to improve the

accuracy of forecasting

Subtasks:
Review current approaches to financial time series forecasting
Compare models and test accuracy
Optimizing parameters of RNN

3/17

Слайд 4

Literature review
Common approaches for analysing financial time series:
1) Classic

statistical methods
Regression models
Autoregressive integrated moving average models
Exponential smoothing
Generalized autoregressive conditionally heteroskedastic methods
2) Artificial neural networks

4/17

Слайд 5

Literature review
Specific features of statistical approaches:
Demonstrate high accuracy result

especially when time series have pattern as trend and/or seasonality
Better work for short-term forecasting
Sensitive to outliers
Optimization of models parameters is quite simple
Do not require much computational power for evaluation

5/17

Слайд 6

Literature review
Specific features of Recurrent Neural Networks:
Able to approximate

complex relationships in time series
Able to forecast for long-term
Optimization of model parameters is quite difficult
Require much computational power for evaluation
Robust to outliers with appropriate parameters' optimization

6/17

Слайд 7

Literature review
Long Short-Term Memory extends the RNN architecture with a

standalone memory

Fig. 1 – Structure of LSTM memory block

7/17

Слайд 8

Literature review
Feature based approach to model selection
8/17

Слайд 9

Literature review
Algorithm of model selection
9/17

Слайд 10

Model comparison
Training details:
Linear Regression
- 91 parameters including bias
Holt-Winters Model
- alpha =

0.55, beta = 0.01, gamma = 0.85
Recurrent Neural Network
- LSTM with 1 layer, with 512 cells.
- The input shape was defined as 1 time series step with 90 features.
- Stochastic gradient descent with fixed learning rate of 0.01 was used as optimizer, loss-function – Mean Squared Error.

10/17

Слайд 11

Model comparison
Data description
Fig. 1 - Daily transaction amount, millions of

rub
(from Nov. 2013 to Apr. 2016 )

11/17

Слайд 12

Model comparison
Results
Fig. 2 – Linear regression
12/17

Слайд 13

Model comparison
Results
Fig. 3 – Holt-Winters Model
13/17

Слайд 14

Model comparison
Results
Fig. 4 – Recurrent Neural Network
14/17

Слайд 15

Model comparison
Results
Table 1 – Performance of models on test data

set

15/17

Слайд 16

Outputs
Findings:
Recurrent Neural Network can outperform the other classical statistical models in

predictive accuracy
More advanced hyper-parameters selection scheme might be embedded in the system to further optimization the learning framework
Selection of model highly depends on time series features

16/17

Optimizing architectures of recurrent neural networks for improving the accuracy of time series forecasts

Содержание

OutlineThe main purpose and subtask of researchLiterature review results Comparison of

The purpose and subtasksThe main purpose:Optimizing RNN parameters to improve the

Literature review Common approaches for analysing financial time series: 1) Classic

Literature review Specific features of statistical approaches: Demonstrate high accuracy result

Literature review Specific features of Recurrent Neural Networks: Able to approximate

Literature review Long Short-Term Memory extends the RNN architecture with a

Literature review Feature based approach to model selection 8/17

Literature review Algorithm of model selection 9/17

Model comparisonTraining details:Linear Regression - 91 parameters including biasHolt-Winters Model - alpha =

Model comparisonData description Fig. 1 - Daily transaction amount, millions of

Model comparisonResults Fig. 2 – Linear regression 12/17

Model comparisonResults Fig. 3 – Holt-Winters Model13/17

Model comparisonResults Fig. 4 – Recurrent Neural Network14/17

Model comparisonResults Table 1 – Performance of models on test data

OutputsFindings:Recurrent Neural Network can outperform the other classical statistical models in

Похожие презентации