4 CONCLUSIONS
In order to make more accurate stock price
predictions, this thesis offers a novel model named
LSTM-GBM, which combines LSTMβs ability to
forecast time series and GBMβs skill to capture and
simulate possibility in the stock market. To assess the
capability of this model, this thesis compared LSTM-
GBM with LSTM model and GBM model. Besides,
this thesis proposed a possible lifting scheme which
is adding a LSTM system after the LSTM-GBM
model. This model is named LSTM-GBM-LSTM.
The performance of LSTM-GBM is also compared
with the performance of LSTM-GBM-LSTM in the
experiment. The results clearly state that LSTM-
GBM is most capable of making predictions among
the four models because LSTM-GBM has made the
most accurate forecast. The MSE of LSTM-GBM is
the lowest while the MAE and the π
ξ¬Ά
score of LSTM-
GBM is closely similar to the result of LSTM. Those
data shows that LSTM-GBM model is able to make
compelling predictions. Therefore, this model might
assist traders and investors to predict stock prices in
stock market. In terms of future works, both LSTM-
GBM and LSTM-GBM-LSTM have the potential to
perform better. For LSTM-GBM, more parameters
are able to be adjusted and more standards of
selecting the target paths are able to be adopted.
Besides, the GBM system might learn more historical
data through some possible improvement. In addition,
the model tends to make predictions of low
expectations due to the design of TPFM and it is able
to make the forcast pricesβ expectation close to the
real data. For LSTM-GBM-GBM, the high MSE and
negative π
ξ¬Ά
score both imply that this model has a
large space for improvement. For example, potential
methods of combining LSTM and GBM are able to
be applied. Those improvements may promote the
ability of the models. Besides, the data may state that
LSTM-GBM-LSTM model has the problem of
overfitting. Therefore, adopting regularization
method or other methods for solving the overfitting
problem may increase the performance of LSTM-
GBM-LSTM model. The problem may be solved by
reducing the complexity of the model as well. A
number of methods are feasible to increase the
performance of LSTM-GBM-LSTM so that the
improvement of this model may be an optional topic
for further studies. In conclusion, LSTM-GBM model
performs the best and it is recommended to adopt this
framework to forecast the share prices in reality.
Meanwhile, greater effort in the future is needed for
reducing the MSE and solving the possible overfitting
problem of LSTM-GBM-LSTM model.
REFERENCES
Ghosh, A., Bose, S., Maji, G., Debnath, N., Sen, S., 2019.
Stock price prediction using LSTM on Indian share
market. Proceedings of 32nd international conference
63, 101-110.
Hao, Y., Gao, Q., 2020. Predicting the trend of stock market
index using the hybrid neural network based on
multiple time scale feature learning. Applied Sciences,
10(11), 3961.
Hodson, T. O., 2022. Root mean square error (RMSE) or
mean absolute error (MAE): When to use them or not.
Geoscientific Model Development Discussions, 2022,
1-10.
Islam, M. R., Nguyen, N., 2020. Comparison of financial
models for stock price prediction. Journal of Risk and
Financial Management, 13(8), 181.
Ji, X., Wang, J., Yan, Z., 2021. A stock price prediction
method based on deep learning technology.
International Journal of Crowd Science, 5(1), 55-72.
Johansson, O., 2022). Stochastic modeling using machine
learning and stochastic differential equations.
Kumar, I., Dogra, K., Utreja, C., Yadav, P., 2018. A
comparative study of supervised machine learning
algorithms for stock market trend prediction. 2018
Second International Conference on Inventive
Communication and Computational Technologies
(ICICCT), 1003-1007.
Lu, W., Li, J., Wang, J., Qin, L., 2021. A CNN-BiLSTM-AM
method for stock price prediction. Neural Computing and
Applications, 33(10), 4741-4753.
Mahesh, B., 2020. Machine learning algorithms-a review.
International Journal of Science and Research (IJSR),
9(1), 381-386.
Nikou, M., Mansourfar, G., Bagherzadeh, J., 2019. Stock
price prediction using DEEP learning algorithm and its
comparison with machine learning algorithms.
Intelligent Systems in Accounting, Finance and
Management, 26(4), 164-174.
Obthong, M., Tantisantiwong, N., Jeamwatthanachai, W.,
Wills, G., 2020. A survey on machine learning for stock
price prediction: Algorithms and techniques. Eprints,
437785.
Rezaei, H., Faaljou, H., Mansourfar, G., 2021. Stock price
prediction using deep learning and frequency
decomposition. Expert Systems with Applications, 169,
114332.
SΓ€rkkΓ€, S., Solin, A., 2019. Applied stochastic differential
equations. Cambridge University Press.
Sherstinsky, A., 2020. Fundamentals of recurrent neural
network (RNN) and long short-term memory (LSTM)
network. Physica D: Nonlinear Phenomena, 404,
132306.
Soni, P., Tewari, Y., Krishnan, D., 2022. Machine learning
approaches in stock price prediction: a systematic
review. Journal of Physics: Conference Series, 2161(1),
012065.
Ta, V. D., Liu, C. M., Tadesse, D. A., 2020. Portfolio
optimization-based stock prediction using long-short
term memory network in quantitative trading. Applied
Sciences, 10(2), 437.