# Multi Armed Bandit and Price Optimization [duplicate]

I obtained a model fit: Sales = 1020 -8.8*Price +2.1*online reputation

after substituting the value of online reputation, I came out with following demand equation: 1036.8-8.8*Price Now to take out the revenue (sales*Price) I converted it to revenue function: 1036.8*Price - 8.8*Price^2

Now can I implement MAB for finding the price in order to maximize the revenue

• Are you sure about using MAB in this particular case? – Adam Przedniczek Jan 13 '17 at 15:54
• No I am not, is it a good idea. If not what shall I use. I m using R – user3713850 Jan 13 '17 at 15:55
• This is the fifth time you have posted the same question. Continuing to do so will likely cause the system to block you from asking any more questions. It would therefore be in your interest to work on clarifying the questions you have posted. – whuber Jan 13 '17 at 17:34
• This time the question is more specific as last time you asked me to keep the question specific – user3713850 Jan 13 '17 at 17:35
• @whuber: my last question was on any optimization applicable. This question is specific on MAB implementations – user3713850 Jan 17 '17 at 13:37

In the general situation $y = ax^2 + bx + c$ the vertex point has coordinates: $(-\frac{b}{2a}, -\frac{\Delta}{4a})$
In your case we're tring to find the vertex of the parabola described by $$y = -8.8 \cdot x^2 + 1036.8 \cdot x$$
Your coefficients are $a=-8.8, b = 1036.8, c = 0$ and $\Delta = b^2 - 4ac = (1036.8)^2$