In this modern age of technology, cyber security is a crucial part. Everything in the current world is run by digital systems and any breach in the system can have catastrophic consequences. Ransomware is one of the most common forms of cyber-attacks. The aim of this research is detecting anomalies in ransomware attacks to stop the attack from happening. In this study we have used classifiers such as: Support Vector Machine (SVM), Naïve Bayes (NV), Random Forest (RF), and Ensemble Model (Stacking Technique) that uses XGboost (XGB), Adaboost (ADB) and Random Forest model together. A simple Artificial Neural Network (ANN) with three hidden layers has also been utilized. The experimentation was conducted on a dataset from Kaggle. The dataset contains 149,043 datapoints and has 14 attributes. In order to preprocess the data, both label encoding and normalization have been applied. Principal Component Analysis (PCA) was used for dimension reduction. K-folding techniques have been used for validation as well. The highest accuracy achieved using an Ensemble Model with the stacking method is 99.35% (standard deviation: 0.0003,k=5), with a 99% confidence interval of [0.992,0.993]. The model mechanism is explained using Local Interpretable Model-agnostic Explanations (LIME), an explainable AI method.