Received: 07 Sep 2025; Accepted: 13 Sep 2025; Published: 07 October 2025

In our thesis, firstly we check if current stock prices are right, looking at the percentage and money differences.

Then, we try to guess what the prices will be tomorrow, comparing real-time and guessed numbers and

explaining the differences. Our suggestions are simple: Sell, Hold, or Buy, to help investors make smart

choices. We also look at what happens if our guesses are wrong and how it affects people’s investments.

In this chapter, the motivation behind the research is introduced. After that, we will present the objectives of

our thesis. This thesis includes the significance of the problem and problem statement in detail. Then we will

present the contributions and significance of the statements. The chapter ends with a short description of the

organization of the thesis.

We need to know how our money in stocks can go up or down, both in percentage and actual value. Trying to

Assess the accuracy of current stock prices.

Determine correctness in percentage and actual value.

Predict next-day stock prices in real-time.

Display and compare predicted and actual prices in percentage and amount.

Provide clear recommendations for buying shares: Sell, Hold, or Buy.

Examine the impact of incorrect share prices on the overall investment portfolio.

We have divided our thesis into five chapters. In Chapter 1, Introduction and some related works are reviewed.

Chapter 2 gives a Literature review that indicates the related work to our thesis. Chapter 3, explains our

research methodology and the mechanism of how we work. Chapter 4, presents an analysis of the results of our

thesis applied to our dataset and discussion. Chapter 5, represents a Conclusion of our thesis.

Filtering an audio signal with an all-pass filter does not usually have a major effect on the signal’s timbre. The

all- pass filter does not change the frequency content of the signal, but only introduces a phase shift or delay.

Audibility of the phase distortion caused by an all-pass filter in a sound reproduction system has been a topic of

many studies, see, e.g., [1], [2]. In this paper, we investigate audio effects processing using high-order all-pass

price analysis around the world. We will go into how we improve the present price accuracy, next-day price

prediction and buying recommendations, and impact on the portfolio.

C. K.

Exploring the intersection of machine learning, quantitative portfolio choice, and mispricing in

financial markets. This abstract highlights the potential of advanced algorithms to identify mispriced assets and

their impact on optimizing portfolio selection strategies, offering valuable insights for investors and researchers

alike. Gu, A., Viens,

F.G. and Yi, B.

considering illiquid assets that are harder to sell. It shows that uncertainty about the duration of illiquidity

increases risk aversion, leading to reduced allocation in both liquid and illiquid assets. Investors are willing to

sacrifice 2% of their wealth to hedge against rare illiquidity crises. Ben-David, I., Drake, M.S. and Roulstone,

D.T.

This study examines how companies make acquisition decisions based on investor perceptions of over

or undervaluation (measured using short interest). Overvalued firms are 54% more likely to acquire other

companies using their stock, while undervalued ones per- form better in cash acquisitions. Misvaluation

influences merger strategies and outcomes. Cvitanic´, J., Lazrak, A., Martellini, L., and Zapatero, F.

differ- ences). Doukas, J.A., Kim, C.F. and Pantzalis, C.

Stocks with higher risk often have bigger price

differences due to the challenges arbitrageurs face in trading them. Sørensen,

C.

Smart investing using stocks and bonds, suggesting zero-coupon bonds for protection. Utilizes

meanvariance approach for optimal wealth growth. Stulz, R.M.

How people from different countries decide

where to invest and how it affects investment returns. It reviews theories, empirical tests, and their significance

in international finance. Davis, C.K.

Why mispricing can occur in the stock market and how it affects

Fig: Methodology Flowchart

The data processing represents evaluating the accuracy of present stock prices by determining correctness in

percentage and amount. It also encompasses predicting next- day prices in real-time, comparing actual and

predicted values. Additionally, it includes processing data to provide practical recommendations for buying

shares at different stages and assessing the impact of incorrect share prices on investment portfolios.

Table 1. Comparison Table

Decision Tree Regression: Used for Predict Current Price Accuracy of a share where we got 98.46% ac-

curacy and 1.54% loss by 70% data for training, 15% data for testing & 15% data for validation.

Random Forest Regression: Used for Predict Cur- rent Price Accuracy of a share where we got 99.01%

accuracy and 0.99% loss by 80% data for training, 10% data for testing & 10% data for validation.

Auto Regressive Integrated Moving Average (ARIMA): Used for Predict Next Day Close Price of a share

where we got 98.19% accuracy and 1.54% loss by 70% data for training, 15% data for testing & 15% data for

validation.

Trading Activity: Volume data represents the number of shares traded.

Choose machine learning models suitable for stock price prediction, such as Random Forest, ARIMA, or

Fig. 1 Predicting Price Accuracy

Fig. 2. Predicting Frequency Accuracy

Fig. 3. Predicting Price Accuracy

Fig. 4. 4 Years Price Accuracy

Moving to Next-Day Prediction Insights, we compare real- time and predicted prices, focusing on differences

Fig. 7. Prices in Candlestick Pattern

Fig. 8. Outcomes

Fig. 9. Equity Curve

Fig. 10. Average Result

Fig. 13. Predicting Time Series

Fig. 14. Predicting Investors’ Portfolio Impaction, delves into the fallout of inaccurate share price pre- dictions

on the investment portfolio.

We looked at predicting stock prices and giving investment advice using machine learning, finding some

important information. Checking how accurate the current prices are shows that the model can tell if they’re

right, measuring correctness in percentages and real amounts. Predicting prices for the next day demonstrated

that the model is good at comparing what it thinks will happen to what re- ally happens, showing the

differences in percentages and amounts. The advice on when to buy or sell, called Strategic Buying