This study investigates the effectiveness of deep learning and classical machine learning techniques in detecting cyberbullying. The research compares five classical machine learning algorithms and three deep learning models. The dataset undergoes preprocessing, including text cleaning, tokenization, stemming, and stop word removal. The performance of the algorithms is evaluated using accuracy, precision, recall, and F1 score metrics. The results show that the proposed technique achieves high accuracy, precision, and F1 scores, with the focal loss algorithm achieving the highest accuracy of 99% and the highest precision of 86.72%. However, recall values were relatively low for most algorithms, indicating difficulty in identifying all relevant data. The study proposes a technique using a convolutional neural network with a bidirectional long short-term memory (Bi-LSTM) layer, trained on a preprocessed dataset of tweets using GloVe word embeddings and the focal loss function. The model achieved high accuracy, precision, and F1 scores, with the GRU algorithm achieving the highest accuracy of 97.0% and the NB algorithm achieving the highest precision of 96.6%. The study highlights the effectiveness of combining pre-trained GloVe embeddings and focal loss in addressing class imbalance and improving performance in cyberbullying detection. The proposed technique demonstrates superior accuracy, precision, and F1 scores compared to traditional machine learning models and other deep learning models. The study also emphasizes the importance of addressing class imbalance in cyberbullying datasets and the potential of deep learning techniques in detecting cyberbullying on social media platforms.This study investigates the effectiveness of deep learning and classical machine learning techniques in detecting cyberbullying. The research compares five classical machine learning algorithms and three deep learning models. The dataset undergoes preprocessing, including text cleaning, tokenization, stemming, and stop word removal. The performance of the algorithms is evaluated using accuracy, precision, recall, and F1 score metrics. The results show that the proposed technique achieves high accuracy, precision, and F1 scores, with the focal loss algorithm achieving the highest accuracy of 99% and the highest precision of 86.72%. However, recall values were relatively low for most algorithms, indicating difficulty in identifying all relevant data. The study proposes a technique using a convolutional neural network with a bidirectional long short-term memory (Bi-LSTM) layer, trained on a preprocessed dataset of tweets using GloVe word embeddings and the focal loss function. The model achieved high accuracy, precision, and F1 scores, with the GRU algorithm achieving the highest accuracy of 97.0% and the NB algorithm achieving the highest precision of 96.6%. The study highlights the effectiveness of combining pre-trained GloVe embeddings and focal loss in addressing class imbalance and improving performance in cyberbullying detection. The proposed technique demonstrates superior accuracy, precision, and F1 scores compared to traditional machine learning models and other deep learning models. The study also emphasizes the importance of addressing class imbalance in cyberbullying datasets and the potential of deep learning techniques in detecting cyberbullying on social media platforms.