About the peak trough trend Segmentation (imaginary method), judging the trend, suddenly come to think of, write down, no experiment. For later use in segmentation

The stock market, constantly undulating, judging how the trend splits up and down,

For example: By looking for the day to draw the trend map, positive turn negative is the crest, negative regularization is the trough

How to distinguish some days are positive, some days negative, even so there can be differentiated up or down trend

Implement Method 1, the crest is calculated (point of time is very important, that is, the division point),

The difference between the peaks and troughs,

Multiple Peaks Direct comparison, if the next value is larger than the previous value is to continue to rise, (a number of falls after the rise?) Is it possible to judge with a drop mean? ）

Drop mean + Crest point turning points (point in time) +

Moving average,

Calculate how long the uptrend has gone,

Pattern recognition .... This hasn't figured out how to identify it like it was before.

About the peak trough trend Segmentation (imaginary method), judging the trend, suddenly come to think of, write down, no experiment. For later use in segmentation