Left ventricular hypertrophy represents the thickened wall condition related to the main pumping chamber of the heart which wither results in blood pressure elevation within the heart or poor action of blood pumping at certain times. The existing methods which contributed to the literature for diagnosing left ventricular hypertrophy using electrocardiogram (ECG) are determined to confirm a comparatively low sensitivity and classification accuracy. Deep learning (DL) approaches which are capable in attaining feature extraction automatically from ECG signals are determined to be potent in detecting cardiac diseases. In this paper, an AlexNet-based deep learning model is proposed for effective and rapid diagnosis of left ventricular hypertrophy using 12 lead ECG. This methodology initially removed the baseline drift and fine noise by denoising signals using discrete wavelet transformation. Then AlexNet model is applied dedicatedly by training and testing 90 and 10% of the data and determining the patterns from the samples and identify unknown patterns. The experimental investigation of the proposed AlexNet-based deep learning model confirmed better sensitivity of 97.21%, superior specificity of 96.79%, and maximized classification accuracy of 99.12%, compared to the CNN-LSTM, CNN-LSTM-AM and ResNet-based left ventricular hypertrophy diagnosis models used for comparison.