Optimization Algorithm in Python
Budget: $30 – $250 USD
Hi, I'm looking for someone to rewrite a better optimization algorithm to find a maximum value over a randomly distribution of outcomes. The current optimization is a hill climbing algorithm and it's not suitable for the random distribution because it looks or a local maximum and stops there.
Here is the problem rephrased.
I have a successive data points (at some time interval) recording the price of a commodity and variables associated with it. At each period, I can decide whether to buy a fixed amount (e.g 100$) or not. You only buy (not selling). I can only buy up to maximum amount (e.g. 1000$) over all the periods.
In each period, the decision rule to buy for 100$ or not is simply taking 4 values of four different data points. The decision rule is simple: buy if all of these values are above a certain threshold.
Applying this decision rule provides a certain “return” at the end of the series, which is equals to the difference between the total amount spent (so a multiple of 100$) and the last price of the commodity multiplied by the amount of it we hold.
I'm looking for a way to find the 4 thresholds that would maximize the return.
I've attached a simplified version of the backtest to show how the return is calculated. I need a script that can change the four values in the file to optimize for the highest return
Here is the problem rephrased.
I have a successive data points (at some time interval) recording the price of a commodity and variables associated with it. At each period, I can decide whether to buy a fixed amount (e.g 100$) or not. You only buy (not selling). I can only buy up to maximum amount (e.g. 1000$) over all the periods.
In each period, the decision rule to buy for 100$ or not is simply taking 4 values of four different data points. The decision rule is simple: buy if all of these values are above a certain threshold.
Applying this decision rule provides a certain “return” at the end of the series, which is equals to the difference between the total amount spent (so a multiple of 100$) and the last price of the commodity multiplied by the amount of it we hold.
I'm looking for a way to find the 4 thresholds that would maximize the return.
I've attached a simplified version of the backtest to show how the return is calculated. I need a script that can change the four values in the file to optimize for the highest return