The latest articles on DEV Community by Vikram Bhagat (@vikram_bhagat_10f70442739). https://dev.to/vikram_bhagat_10f70442739 https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F1890157%2F5e7e03c2-9f34-4496-b0c8-6be73c567e24.png https://dev.to/vikram_bhagat_10f70442739 en Vikram Bhagat Tue, 06 Aug 2024 09:00:25 +0000 https://dev.to/vikram_bhagat_10f70442739/how-to-improve-ml-model-accuracy-for-text-classification-4l28 https://dev.to/vikram_bhagat_10f70442739/how-to-improve-ml-model-accuracy-for-text-classification-4l28 <p>Hi Experts,</p> <p>We are dealing with Text Classification Problem. We have around 80K records with around 50 classes. The data is highly imbalanced. It has 2 columns one for description and other contains class.<br> Till now we have tried following models and techniques:</p> <ol> <li>Data Preprocessing: a. Lowercase conversion, removed numeric texts, removed punctuations b. Removed unimportant words and stop words c. Lemmatization</li> <li>TFIDF transformation</li> <li>Using SKLEARN Models: a. Linear SVC b. Linear Regression c. Logistic Regression d. Decision Trees e. Random Forest</li> <li>Using Huggingface Transformers: a. Google Bert b. Distil Bert</li> <li>SMOTE sampling</li> </ol> <p>It is observed that the maximum accuracy we got is 70% (Random Forest and Google Bert).<br> Is there any scope to improve accuracy?<br> If yes, what other techniques or models we can use to improve accuracy?</p> machinelearning datascience python scikitlearn