I came across a question in a quiz for my son as he asked for help. Well I went through classic way and got a mixture of exponential and quadratic expression, I derived to look for the extremum then moved to second derivative. But I felt I was locked in the same problem. so we have $ P(x)= x^2e^x -e^x -1 >0$
solving for P'(x)=0 to look for extremums, we have $e^x(x^2 +2x-1)= 0 \rightarrow x_{1,2}=\frac{-2\mp \sqrt8}{2} $
So $x_{1,2}=\frac{-2\mp \sqrt8}{2} $ are P'(x) roots
The probem is that the graph plotting gives differents roots. As I wonder how shoudl I proceeed, I had thought to Newton method. But is it the only way to do so . Thanks for any help.
If you look for the zero of function $$f(x)=e^x(x^2-1)-1$$ a simple inspection shows that the solution is greater than $1$.
Make a series expansion $$f(x)=-1+2 e (x-1)+3 e (x-1)^2+2 e (x-1)^3+\frac{5}{6} e (x-1)^4+O\left((x-1)^5\right)$$ Now, series reversion $$x=1+\frac{f(x)+1}{2 e}-\frac{3 (f(x)+1)^2}{8 e^2}+\frac{7 (f(x)+1)^3}{16 e^3}-\frac{235 (f(x)+1)^4}{384 e^4}+O\left((f(x)+1)^5\right)$$ SInce you want $f(x)=0$, an estimate is $$x=1+\frac{1}{2 e}-\frac{3 }{8 e^2}+\frac{7 }{16 e^3}-\frac{235 }{384 e^4}=1.14376$$ while the solution given by Newton method is $1.14776$.