The goal of the present article is extending and developing econometrics and network structure based methods which are able to distinguish PRICE manipulation in Tehran STOCK exchange. The principal goal of the present study is to offer model for approximating PRICE manipulation in Tehran STOCK exchange. In order to do so by applying separation method a sample consisting of 397 companies accepted at Tehran STOCK exchange were selected and information related to their PRICE and volume of trades during years 2001 until 2009 were collected and then through performing runs test, skewness test and duration correlative test the selected companies were divided into 2 sets of manipulated and non manipulated companies. In the next stage by investigating cumulative return process and volume of trades in manipulated companies, the date of starting PRICE manipulation was specified and in this way the logit model, artificial neural network, and by using information related to size of company, clarity of information, ratio of PIE and liquidity of STOCK one year prior PRICE manipulation; a model for FORECASTING PRICE manipulation of STOCKs of companies present in Tehran STOCK exchange were designed. At the end the power of FORECASTING models were studied by using data of test set. Whereas the power of FORECASTING logit model for test set was 92.1 %, and for artificial neural network was 94.1%; therefore both of the 2 aforesaid models has high power to anticipate PRICE manipulation and there is no considerable difference between FORECASTING power of these 3 models.

The goal of the present article is extending and developing Multiple Disciminant Analysis (MDA) method which is able to distinguish buble PRICE in Tehran STOCK exchange. The principal goal of the present study is to offer model for approximating buble PRICE and also the factors efficient to make the model work at Tehran STOCK exchange. In order to do so by applying separation method a sample consisting of 397 companies accepted at Tehran STOCK exchange were selected and information related to their PRICE and volume of trades during years 2001 until 2009 were collected and then through performing runs test, skewness test and duration correlative test the selected companies were divided into 2 sets of with bubble PRICE and non bubbled companies. In the next stage by investigating cumulative return process and volume of trades in bubbledted companies, the date of starting bubble PRICE was specified and in this way the multiple discriminant analysis, and by using information related to size of company, clarity of information, ratio of P/E and liquidity of STOCK one year prior bubble PRICE; a model for FORECASTING bubble PRICE of STOCKs of companies present in Tehran STOCK exchange were designed. At the end the power of FORECASTING model was studied by using data of test set. Whereas the power of FORECASTING MDA model was 90.2%; the model has high power to anticipate bubble PRICE.

FORECASTING STOCK market PRICE index has always been a challenging task, since it is affected by many economic and non-economic factors; therefore, selecting the best and the most efficient FORECASTING model is difficult. The time series in the real world, including the STOCK PRICE index time series, rarely have a pure linear or non-linear structure. The Exponential Smoothing Model, Autoregressive Integrated Moving Average Model, and Nonlinear Autoregressive Neural Network can be used to make forecasts based on time series. In this research, to take advantage of all these models and to reduce FORECASTING errors, a novel approach was tested by the linear combination of the results of these models. Weights used to combine the results, were determined using Genetic Algorithm and also equal weights. After determining the predictability of time series (using variance ratio test) the proposed hybrid methods were used on a monthly set of Tehran STOCK Exchange PRICE Index (TEPIX). The results showed an improvement in forecasts made by this method with using equal weights compared to each of its constituent models.

FORECASTING financial markets is an important issue in finance area and research studies. Importance of FORECASTING on one hand and its complexity, on the other hand, researchers have done much work in this area and proposed many methods. In this research, we propose a hybrid model include wavelet transform, ARMA-EGARCH and NN for day-ahead FORECASTING of STOCK market PRICE in different markets. At first WT is used to decompose and reconstruct time series into detailed and approximated parts. And then we used ARMA-EGARCH and NN models respectively for FORECASTING details and approximate series. In this model we used technical index by approximate part to the improvement of our NN model. Finally, we combine prediction of each model together. For validation, proposed model compare with ANN, ARIMA-GARCH and ARIMA-ANN models for FORECASTING STOCKs PRICE in UA and Iran markets. Our results indicate that proposed model has better performance than others model in both markets.

Time series processes can be classified to three models, linear models, stochastic models and chaotic models. Based on these classification the linear models are forecastable, the stochastic models are unforecastable and the chaotic models are semi forecastaable. The previouse researches in the modeling and FORECASTING of the STOCK PRICE usually try to prove that, the fluctuations of the share PRICEs in Tehran STOCK Exchange are not random walks in spite of the existance similarity to the random walks. Indeed the market has a chaotic behavior. This means that, the Efficient Market Hypothesis (EMH) is failed. Therefore by using a complex and powerfull models such as artificial neural networks, one can forecast STOCK PRICEs in tehran STOCK merket. This paper proposed another approach to modeling and FORECASTING of the share PRICE. This approach is based on the Stochastic Differential Equations. The modeling is based on the Black- Scholes pricing model. Comparison the simulation result with the linear ARIMA model, indicates that the proposed structrure, provides an accurate next step and the long term share PRICEs and daily returns FORECASTING.

STOCK PRICE FORECASTING is very important in STOCK exchange, because it indicates desirable or undesirable investing condition for investors and STOCKholders. Among FORECASTING methods, neural networks and fuzzy logic were used in many areas and have some advantages and disadvantages. So, this paper intends to combine fuzzy reasoning theory with neural networks in order to improve the precision and convergence speed of FORECASTING model. Thus, a neural network-driven fuzzy reasoning system is proposed on the basis of improved Takagi-Sugeno reasoning model.In this paper, a neuro-fuzzy approach was employed to forecast STOCK PRICE and it was compared with ARIMA model based on six statistical parameters. The experimental results showed that the fuzzy neural networks based on six indices measurement was better than ARIMA model and has such properties as fast convergence, high precision and strong function approximation ability and it is suitable for real STOCK PRICE FORECASTING.

The study of combining FORECASTING in Tehran STOCK exchange has been discussed in this survey. Combining FORECASTING is a new method and various studies have been done about this subject the results of which have shown most reduction FORECASTING errors. In this study multiple regression model is used to combine individual FORECASTING methods. To do so, several FORECASTING methods were studied and five methods (Box - Jenkins, linear smoothing, Holt, Power & quadratic trend) were selected. Pars Electric company was subjcted to a case study with 159 weekly data. These five methods, were evaluated based on multiple regression with stepwise method and finally, two methods (Box - jenkins and quadratic trend) adjusted. The errors of this model were compared with the other used methods (MAD, MAPE, MSE). The errors rate of combining FORECASTING methods were compared with the best FORECASTING method. The result showed that MSE ,MAD, and MAPE of combining model were 0.064, 0.26, and 0.24 to the best method respectively. These indicate considrable reduction in the error of combining FORECASTING comparing to the other methods

STOCK market is affected by news and information. If the STOCK market is not efficient, the reaction of STOCK PRICE to news and information will place the STOCK market in overreaction and under-reaction states. Many models have been already presented by using different tools and techniques to forecast the STOCK market behavior. In this study, the reaction of STOCK PRICE in the STOCK market was modeled by the behavioral finance approach. The population of this study included the companies listed on the Tehran STOCK Exchange. In order to forecast the STOCK PRICE, the final PRICE data of the end December, March, June, and September 2006-2015 and the STOCK PRICEs of 2014 and 2015 were analyzed as the sample. In this study, Bayes' rule was used to estimate the probability of the model change. Through this rule, the probability of an event can be calculated by conditioning the occurrence or lack of occurrence of another event. The results of model estimation showed that there is the probability of being placed in high-fluctuated regimes (overreaction) and low-fluctuated (under-reaction of STOCK PRICE despite the shocks entered to the STOCK market. In modelling with the 4-month final PRICEs, it was proved that the real STOCK PRICE had no difference from the market PRICE.

The capital market plays a crucial role within a country's financial structure and is instrumental in funding significant, long-term projects. Investments in the railway transport industry are vital for boosting other economic areas and have a profound impact on macroeconomic dynamics. Nonetheless, the potential for delayed or uncertain returns may deter investors. Accurate predictions of rail company STOCK PRICEs on exchanges are therefore vital for making informed investment choices and securing sustained investment. This study employs deep learning techniques to forecast the closing PRICEs of MAPNA and Toucaril shares on the Tehran STOCK Exchange. It utilizes deep neural networks, specifically One-dimensional Convolutional Neural Networks (1D-CNN), Long Short-Term Memory (LSTM) networks, and a combined CNN-LSTM model, for STOCK PRICE prediction. The effectiveness of these models is measured using various metrics, including MAE, MSE, RMSE, MAPE, and R2. Findings indicate that deep learning methods can predict STOCK PRICEs effectively, with the CNN-LSTM model outperforming others in this research. According to the results, The CNN-LSTM model reached the highest R2 of 0.992. Also, based on criteria such as MAE, MSE, RMSE, and MAPE the best results belong to LSTM (Kaggle-modified) with 521.715, 651119.194, 806.920, and 0.028, respectively.

STOCK exchange is a secure way to gain public trust for investment in different securities with varying risks. In this way small and scattered capitals which cannot be utilized alone can be accumulated and a huge investment can be made of them for economic development and progress. In STOCK exchange, there are a lot of sensitivities to PRICE formation course. This has caused that changes related to such phenomenon to be systematically analyzed. In recent years, a variety of models have been employed by specialists for prediction of share PRICE. Since Artificial Intelligence Techniques which include Neural Networks, Genetic Algorithm and Fuzzy Logic have achieved successful results in complex problems, they are used more for this purpose.Present study intends to answer the question whether using a combination of Artificial Intelligence Technique, a model can be set up which compared to other linear and non-linear methods predicts share PRICE with less error. In this research, to predict STOCK PRICE (Tehran STOCK exchange - Iran Khodro CO), a combination of Artificial Intelligence Methods including Neural Networks, Fuzzy Logic and Genetic Algorithm are used and this combined model is compared with Neural Network Methods, the title for one of the other Artificial Intelligence Models, and ARIMA linear model, given R2, MAE, MAPE, MSE. The results of this research shows that the superiority of Hybrid model compared to other models are examined.