Test

After hitting golf balls with our fairway wood golf club on the driving range, we got data and we need to select the relevant ones.

So we have decided to keep and build our dataset with this :

• ball Carry
• ball Speed
• ball Apex
• ball Launch Angle
• ball Dispersion

We have a dataset which was cleaned and normalized with Python process. This dataset now looks like this for balls hit by a our fairway wood golf club: 

We have this scatter plot for 10 golf shots. Of course, we could do it with 100 or 1000 golf shots but time, cost and light bulbs in hands were our limits:

In order to answers the questions related of our part 1 article, we wanted to draw correlation between these data. In golf, we want to improve distance and accuracy. So we have built a Heat Map in order to detect and define the correlation between each caracteristics. A Heap Map is is a graphical representation of data where values are depicted by color.

Regarding our dataset, the more the color is white, the stronger the correlation between caracteristics of our data set is high.

Here, we can observe that the Carry distance (meters) is hightly correlated to Ball Speed and Launch Angle.

We can notice also that the Carry distance is less correlated to the Apex (how high fly the ball).

Then, it helps us to decide which features of the dataset is important. For this, we had a limit of a correlation coefficient of 0,4. It means that we exclude all features where correlation coefficient is not strong enough to be relevant.

L’article Golf: part 2 – use of Machine Learning with Trackman Range est apparu en premier sur Golf and technologies : machine learning, artificial intelligence, mobile app.

We can wonder a couple of questions about golf sport and data we start to gather as amateur or non-professional golfer:

• how can I hit my Driver or golf clubs longer?
• which are the parameters that can bring me more power to hit the golf ball?
• is there any relationship between my golf ball flight angle and the distance I can do?
• is what says my golf instructor is based on numbers, on facts or on intuitive observations?
• more generally, how the data I collect with Trackman (or any golf ball tracker) can be analysed?

In order to answer these questions, we need a couple of tools:

A driving range equiped with a golf ball tracker – TrackMan Driving Range will be perfect

Golf clubs and balls of course

A computer to execute machine learning models via Python language and Jupyter code execution tool helper

In order to give some beginning of answers of our questions, we have decided to use ML or Machine Learning. We can remember here that Machine learning is a branch of artificial intelligence (AI) and computer science which focuses on the use of data and algorithms to imitate the way that humans learn, gradually improving its accuracy.

L’article Golf: part 1 – use of Machine Learning with Trackman Range est apparu en premier sur Golf and technologies : machine learning, artificial intelligence, mobile app.