We have a study that uses previous data of Mortality $($number of deaths$),$ Feeds Consumption, and Average Live Weight from a Chicken Farm. Using time series forecasting, we are then to predict these $3$ once more $($Mortality$,$ Feeds Consumption, and Average Live Weight$)$ and put it side by side with the actual collected data from the same day. The question is regarding another feature which aims to detect anomalies in the data. For example, if the actual average live weight is lower than the predicted then that counts as an anomaly. The question is what anomaly detection algorithm should be used for this case, and if none then what other means can we do$.$ Thanks