Urinary Infection represents a prevalent issue globally leading numerous individuals to seek urgent medical attention. Diagnosing urinary infection stones preoperatively poses a challenge and accurately determining stone composition is typically only feasible ex vivo. To increase the perioperative management along with postoperative precaution of infection stones which comes up with a ML technique for preoperatively acknowledging the infection stones in vivo. Amid 2565 individuals included where 1168 qualified individuals with urinary calculi were randomly split into training (75%) and test (25%) sets. The forecast technique was developed by utilizing two ML methods and 14 preoperative factors and it’s carrying out was judged by calculating the area under the ROC of the validation set. This study analyzed the significance of the 14 variables in each prediction technique for predicting infection stones. The validation set comprised 89 individuals with infection stones. The two output methods exhibited a strong bias in the validation set (AUC: 0.78 & 0.77). The LASSO technique was chosen as the final technique. Urine culture positivity and urine pH emerged as two main evocator of infection stones. Through machine learning this study build a predictive technique which is capable of promptly identifying infection stones in vivo with noticeable predictive execution. This technique could facilitate hazard assessment as well as decision-making aid for infection stones which thereby optimizes disease handling of urinary calculi as well as improving patient prognosis.