Snow avalanche is an inevitable issue that is faced by mankind residing near the hilly and the mountainous regions. It is a natural disaster that is frequently observed in such terrains. Prediction of these avalanches is crucial for wellbeing of mankind. The concept of using cosine similarity with nearest neighbour is an innovative idea in nearest neighbour based avalanche forecasting model. The results of the model are encouraging, but a need for a mechanism that will provide additional preference to the significant parameters is observed. Present work focuses on the application of weighting factor to the nearest neighbour model with cosine similarity. Use of weighting factor helps in further tuning of the forecasting model. Selection of weighting factors for each parameter is accomplished by considering the effect of each parameter on the avalanche activity. The accuracy of the model is gauged using performance measures - Critical Success Index and Bias and by the changes reflected in the confusion matrix. An increase of 0.1978 and 0.4167 is observed in the values of Critical Success Index after the application of the weights to the forecasting model for dataset combination I and II respectively. The proposed work is implemented using the snow and meteorological data for the Bahang region of Himachal Pradesh, India.