In the field of cable insulation evaluation, there are problems such as low evaluation accuracy and difficulty in quantitatively judging the insulation status. This paper proposes an integrated algorithm approach based on four commonly used machine learning algorithms: Support Vector Machine (SVM), Extreme Gradient Boosting (XGBoost), Random Forest (RF), and K-Nearest Neighbors (KNN) for comprehensive evaluation of cable insulation conditions. The Bayesian optimization algorithm is employed to tune the hyperparameters of SVM and XGBoost. Then, the operating years, partial discharge, and visual conditions in the Canadian cable dataset are used as feature inputs. The calculated results indicate that the accuracy, precision, and recall of the four models in the test set all reached over 96%, and they have good performance. Following the integration of the four algorithms, the evaluation accuracy and stability are further enhanced. The multi-algorithm ensemble evaluation method proposed in this paper can accurately and effectively assess the insulation status of cables, and provide reference for the relevant personnel.