Trend surfing

[chrusso]

Hi guys, looking for some help on thistrying to think about a proper algorithm to anticipate a bull run, or, in other words, I'm looking to make a small PHP program that allows to buy a coin right after the prices go up.

Now, I've been trying to get some patterns that could indicate that a certain coin is going up, I've been trying to find correlations. Monitoring a few websites in order to fetch and parse the news could be useful.

There's also a few open source projects that allow to parse emotions on a given text, for example:

https://github.com/Chaiavi/TextToEmotion

TextToEmotions reads the block of text, parses emotions from it, then returns a response object containing the percent value probability of each emotion.

Or here:

http://krcadinac.com/synesketch/

Any ideas about any other indicators that could shortly anticipate an uptrend?

Thanks!

[1714987886]

1714987886

[1714987886]

1714987886

[1714987886]

1714987886

[hase0278]

Any ideas about any other indicators that could shortly anticipate an uptrend?

Thanks!

Other indicators you can use to anticipate an uptrend is by using technical analysis. IMO this type of analysis is good when paired with Fundamental analysis(the one you are attempting to use since you'll make the program predict a bull run using market sentiment). By utilizing the two, somehow it can make that small PHP program more accurate in predicting an uptrend.

[chrusso]

Any ideas about any other indicators that could shortly anticipate an uptrend?

Thanks!

Other indicators you can use to anticipate an uptrend is by using technical analysis. IMO this type of analysis is good when paired with Fundamental analysis(the one you are attempting to use since you'll make the program predict a bull run using market sentiment). By utilizing the two, somehow it can make that small PHP program more accurate in predicting an uptrend.

Hi hase0278, thanks for your feedback!

Do you know anything about technical indicators that we can use in order to increase the performance?

Using the Poloniex API, for example, we can get the historical price, which I believe would allow us to calculate the variables we need, here's an example:

https://poloniex.com/public?command=returnChartData&currencyPair=BTC_ETH&start=1405699200&end=9999999999&period=14400


- currencyPair=BTC_ETH <defines the pair that we want to lookup>
- start=1405699200 <defines the last date we want to lookup>
- end=9999999999 <defines the last date we want to lookup, using 9999999999 we get all the records until present day>
- period=300 <defines the interval of each candle in seconds, set in 300 groups the information every 300 seconds (5 minutes)>

The first question would be how many days we would need to capture in order to perform a proper technical analysis?

Would it be alright calculating by the last 90 days? If so, how long would you define the interval? 5 minutes? half an hour? 1 hour? 12 hours? a day?

Can anyone recommend how to perform a proper technical anaylsis from this data?

Thanks!!
Kind regards;
Chris

[luckyluigi]

Oh, I would like to know for sure what will happen and when, but here I`m not skilled, so just depending on a wind of changes.

[chrusso]

I just wanna detect when the wind start to blow stronger, and buy with an stop loss just in case.