Stock Market Price Prediction and Data Analysis -- 2
Budget: $250 – $750 USD
Using data from historical data and multiple (3-5) forecast websites and a manually entered source, try to predict the next day Max/Min value and the aproximate low/high peak times of any Stock Symbol. Not all the sources provide forecast for all the Stock Symbols.
Should be able to “grab” next day Max/Min prediction data from websites right before the market closes. Each stock may have different closing time.
Will be multiusers (5) and a "assistent" user, only allowed to enter manually one of the forcast sources.
Each user will have its own list of stock symbol (20-30). Each user will add or archived Stock Symbols as needed. Only those Stock Symbols will be tracked.
Should record, per day, what was the predicted Max/Min/peak info per source per Stock Symbol, and, after the market closes, what was the actual Max/Min/peak for each Stock Symbol.
The forecasts and historical data are only a piece of information. Doing data analysis, you have to come with your own and more accurate forecast for the next day.
This is the way we’re doing it now. (Video.mp4). My assistant send me every day the daily graph of each Stock Symbol I’m investing on/or reviewing.
This is the forecast of one of the sources for next day. (Picture 01.jpeg). They give up to 14 days of forecast, but the closer the date the more accurate.
Ono of the other source is this. (Picture 02.jpeg). They provide up to 30 days of forecast. Same thing, the closer the more accurate.
Depending the symbol, it will be a different “market hours”. You have to check the forecast within the last 15 minutes of the market day.
I need, on daily basis, the daily graph of previous day and the forecasts, and be able to “Slide” the graphs to see the behavior of previous days as shown in Video.mp4.
This was "the day before" graph (RED rectangle) and "today’s" forecast (Info at the top of picture) (Picture 03.jpeg & Picture 04.jpeg)
Once the new day stars all sources erases "today’s" forecast. That’s why it’s so important to take the forecast before the end of the “market day”.
I need to register:
- Openning value
- Minimum Value
- Maximum Value
- Closing value
- Low Peak time(s)
- High Peak time(s)
- The graph
And from there calculate the flotation band which is the difference between the Max/Min. (Picture 05.jpeg). In this case, the flotation band was $0.45
And register time of peeks (Low/High) within the top and bottom 15% of the flotation band. (Customizable per stock symbol) (Please see Picture 06.jpeg). 15% of $0.45 = $0.0675.
Anytime the value goes inside the top and bottom 15% (customizable for each stock symbol) it’s a peek. Like in this graph (Picture 06.jpeg), there can be multiple peeks a day. You have to register the times at which the peeks occurred.
Every time the value of any symbol enter in the top and bottom peek zone, a push notification have to show in the iPhone screen advising what symbol is in a peek zone, and an audible alarm will sound EVEN if the iPhone is in vibrate mode.
The core of the software can be a computer or web-based but the dashboard and main interface have to be iOS or mobile friendly.
Please anticipate what range of accurancy you'll be able to provide. Before doing all the work, and without knowing what formulas, algorithms, tools, etc. you will use to provide forecast, a 15 days test period will be needed to measure acurrancy.
Should be able to “grab” next day Max/Min prediction data from websites right before the market closes. Each stock may have different closing time.
Will be multiusers (5) and a "assistent" user, only allowed to enter manually one of the forcast sources.
Each user will have its own list of stock symbol (20-30). Each user will add or archived Stock Symbols as needed. Only those Stock Symbols will be tracked.
Should record, per day, what was the predicted Max/Min/peak info per source per Stock Symbol, and, after the market closes, what was the actual Max/Min/peak for each Stock Symbol.
The forecasts and historical data are only a piece of information. Doing data analysis, you have to come with your own and more accurate forecast for the next day.
This is the way we’re doing it now. (Video.mp4). My assistant send me every day the daily graph of each Stock Symbol I’m investing on/or reviewing.
This is the forecast of one of the sources for next day. (Picture 01.jpeg). They give up to 14 days of forecast, but the closer the date the more accurate.
Ono of the other source is this. (Picture 02.jpeg). They provide up to 30 days of forecast. Same thing, the closer the more accurate.
Depending the symbol, it will be a different “market hours”. You have to check the forecast within the last 15 minutes of the market day.
I need, on daily basis, the daily graph of previous day and the forecasts, and be able to “Slide” the graphs to see the behavior of previous days as shown in Video.mp4.
This was "the day before" graph (RED rectangle) and "today’s" forecast (Info at the top of picture) (Picture 03.jpeg & Picture 04.jpeg)
Once the new day stars all sources erases "today’s" forecast. That’s why it’s so important to take the forecast before the end of the “market day”.
I need to register:
- Openning value
- Minimum Value
- Maximum Value
- Closing value
- Low Peak time(s)
- High Peak time(s)
- The graph
And from there calculate the flotation band which is the difference between the Max/Min. (Picture 05.jpeg). In this case, the flotation band was $0.45
And register time of peeks (Low/High) within the top and bottom 15% of the flotation band. (Customizable per stock symbol) (Please see Picture 06.jpeg). 15% of $0.45 = $0.0675.
Anytime the value goes inside the top and bottom 15% (customizable for each stock symbol) it’s a peek. Like in this graph (Picture 06.jpeg), there can be multiple peeks a day. You have to register the times at which the peeks occurred.
Every time the value of any symbol enter in the top and bottom peek zone, a push notification have to show in the iPhone screen advising what symbol is in a peek zone, and an audible alarm will sound EVEN if the iPhone is in vibrate mode.
The core of the software can be a computer or web-based but the dashboard and main interface have to be iOS or mobile friendly.
Please anticipate what range of accurancy you'll be able to provide. Before doing all the work, and without knowing what formulas, algorithms, tools, etc. you will use to provide forecast, a 15 days test period will be needed to measure acurrancy.