Mathfi-Sports
A new breed of next generation AI-powered service, helping your sport platform or human experts to make decisions by performing binary or multi-class predictions and classifications or sanity check their existing predictions/classifications. Some examples:
1. Decide whether a football player will churn and be transferred in next summer: Knowing 1000s of past football players’ data features such as name of the player, club name, age in that season, position, appearances, matches played, goals scored, yellow cards per game, red cards per game, shots per game, passes success (%), aerials won per game, man of the match award, global rating, tackles per game, interceptions per game, fouls per game, offsides won per game, times dribbled per game, own goal per game, key passes per game, dribbling per game, times fouled per game, offsides per game, times disposed per game, bad controls per game, average passes per game, and many more other data features, and knowing whether or not that player was transferred (target label of prediction), predict whether a current season football player will be transferred during next summer.(and with what probability).
2. Predict if a sport club/team will win a future match: Considering the collective feature data of all players of sport teams/clubs of (both sides of) 1000s of past sport matches (e.g. football) including name of the players, club name, age in that season, position, appearances, matches played, goals scored, yellow cards per game, red cards per game, shots per game, passes success (%), aerials won per game, man of the match award, global rating, tackles per game, interceptions per game, fouls per game, and if available exercise hours or types per day, body sensor data, heartbeat, stress test data, age, gender, diet, heartbeat, health history, height, BMI, weight, blood pressure, urine samples, etc, and knowing which one of those two teams have won the past match, decide which one of the two new teams involved in a new match will be the winners? (and with what probability).
3. Predict the winner of the future NBA playoffs: Which one will win? home team or away team: Having the past statistics of NBA (say for example statistics of couple of the most recent seasons), including the unique ID for the game, the name of the home team, the number of points scored by the home team, the field goal percentage of the home team, the free throw percentage of the home team, the three-point field goal percentage of the home team, the name of the away team, the number of points scored by the away team, the field goal percentage of the away team, the free throw percentage of the away team, the three-point field goal percentage of the away team and knowing that which one was the winner (1: Home Team Win or 0: Away Team Win, label and target of the prediction), for current season predict whether home team will be the winner or the away team (and with what probability).
4. Decide whether to hire an athlete or not: Knowing the past athletes’ data features such as exercise hours or types per day, body sensor data, heartbeat, stress test data, age, gender, diet, heartbeat, health history, height, BMI, weight, blood pressure, urine samples, etc. and knowing what target performance they have achieved in past for example a speed class of (A, B, C or D) in a race, classify which one of the performance levels (A, B, C or D) a new candidate athlete will achieve in a future race in order to decide whether to hire/select them or not. (and with what probability).
5. Decide whether an athlete is doping: Having the past athletes’ data features including urine sample, blood sample, body sensor data, heartbeat, urine samples, stress test data, blood pressure, exercise data, etc, and knowing whether those athletes have doped (using illegal substances) in certain pint of time, decide whether a new athlete is doping? (and with what probability).
and any other binary or multi-class sport classification or prediction use cases.