I am so frustrated with my final year project..this is literally the task I am asked to do

Wow, which school and course is that

You have to use time-series forecasting, since the stock-market fluctuations essentially behave as a time series depending on (daily, weekly, monthly or annual) seasonal trends (per industry or market/stock type). You should however be careful because sometimes rumors in the market can cause the time-series to behave unpredictably, rendering your projections useless. Inbox me for more info - cmdimkpa@gmail.com

montydimkpa:
You have to use time-series forecasting, since the stock-market fluctuations essentially behave as a time series depending on (daily, weekly, monthly or annual) seasonal trends (per industry or market/stock type). You should however be careful because sometimes rumors in the market can cause the time-series to behave unpredictably, rendering your projections useless. Inbox me for more info - cmdimkpa@gmail.com

Gosh, can we please stop this inbox me for more 'obvious' info. It just kills the real reason of a 'forum'

Back to topic

There is absolutely no way you can build such system without manual influence of the prediction results.
Probably will work as the football betting systems work

cnjy:
I am so frustrated with my final year project..this is literally the task I am asked to do

You need to implement "Random Walk Hypothesis", that's a tough job, but it's not impossible.

Just an idea

Urine:


You need to implement "Random Walk Hypothesis", that's a tough job, but it's not impossible.

Tough job?
Randomly selecting numbers out of previous records?

Or am I missing something?

spikesC:


Tough job?
Randomly selecting numbers out of previous records?

Or am I missing something?

Bro, you have to realise some of us are not as smart as you.

Urine:


Bro, you have to realise some of us are not as smart as you.

Pls don't get me wrong, I was actually thinking I don't understand the Random Walk Hypothesis implementation.
So am trying to make you explain more (of the implementation )

spikesC:


Pls don't get me wrong, I was actually thinking I don't understand the Random Walk Hypothesis implementation.
So am trying to make you explain more (of the implementation )

Lool. I was just joking, RW is the mathematical formalization of a process that consists of successive random steps. It's used to model stock prices in economics.

Urine:


Lool. I was just joking, RW is the mathematical formalization of a process that consists of successive random steps. It's used to model stock prices in economics.

From the little research I did, it showed that the theory doesn't have a mathematical formula but rather implies that the output data can never be predicted but can only be a random value from the previous output values.

This implies that, if the OP can actually debate it with the lecturer, he would be allowed to develop a system that randomly selects a value from previous stock-market data which was fed into the system.

This randomness can certainly be mitigated by developing an algorithm that predicts the output by current market fluctuations (which stock is performing well), stock history value (how the stocks has been performing), previous time frame performance (how the stock performed in that particular time of every year) and other relational data. But then, it definitely won't be perfect

spikesC:


From the little research I did, it showed that the theory doesn't have a mathematical formula but rather implies that the output data can never be predicted but can only be a random value from the previous output values.

This implies that, if the OP can actually debate it with the lecturer, he would be allowed to develop a system that randomly selects a value from previous stock-market data which was fed into the system.

This randomness can certainly be mitigated by developing an algorithm that predicts the output by current market fluctuations (which stock is performing well), stock history value (how the stocks has been performing), previous time frame performance (how the stock performed in that particular time of every year) and other relational data. But then, it definitely won't be perfect

You see why I say it's a tough job. I like the way you presented your thoughts, people like you make me feel I still have a lot to learn. Well done!

Urine:


You see why I say it's a tough job. I like the way you presented your thoughts, people like you make me feel I still have a lot to learn. Well done!

Pls don't flatter me oo, we all do have a lot to learn

all yu need is A.P.I bro..nd more d app is already on windows mobile store ...

When I saw your post, I just concluded that u haven't done any serious Googling. There are tons of tutorials out there on the Internet on Stock Market Analysis.

@OP,i dont know if you are IT inclined or just an economics student as i would have suggested you used a machine learning algorithm to train your software in order to carry out future predictions(regression problem).If this is a bit out of your case study then you need not bother yourself about my suggestion and just find a suitable alternative.#my2naira

cnjy:
I am so frustrated with my final year project..this is literally the task I am asked to do

Without stress, you will need to contact Stock Brokers or Agents and inquire how they perfect their predictions. It is based on this analysis that you will now work upon. After drawing the Diagram chart or flow chart, you will now think of the programming language and the other materials you would need to build a robust technology on this.
IT broadens your scope of thinking and imagination.