Page 165 - Web_Application_v2.0_C12_Fb
P. 165

  Manufacturing and Supply Chain: ML is used to optimise production processes, predict maintenance
                   needs, and streamline supply chain management. Predictive maintenance systems can foresee equipment
                   failures, reducing downtime and maintenance costs.
                   Gaming: In the gaming industry, ML is used to develop AI that can learn and adapt to player behaviour,
                   making games more challenging and  engaging.  For instance, reinforcement learning is used in video
                   games like AlphaGo to improve game strategies.

                 7 Basic Steps in Machine Learning
                 Different steps in Machine Learning are as follows:
                 1. Data Collection: Acquire relevant data from various sources, ensuring it is sufficient and appropriate for
                   the problem being addressed.

                 2. Data  Preparation: Process  and clean the collected data by handling missing  values, outliers, and
                   inconsistencies, and format it to ensure it is ready for model training.
                 3. Model Selection: Select the most suitable machine learning algorithm or model based on the task (e.g.,
                   classification, regression, clustering) and the data's characteristics.
                 4. Model Training: Train the chosen model using the prepared data, enabling it to identify and learn patterns
                   and relationships.
                 5. Model Evaluation: Evaluate the model’s performance using relevant metrics like accuracy, precision, recall,
                   or F1-score to determine how well it performs.
                 6. Parameter Tuning: Fine-tune the model by adjusting hyperparameters to optimise its performance and
                   achieve better results.

                 7. Prediction: Apply the trained and optimised model to make predictions or decisions on new, unseen
                   data.

                                 Prediction

                                         Parameter Tuning



                                                   Model Evaluation


                                                               Model Training



                                                                          Model Selection


                                                                                   Data Preparation



                                                                                              Data Collection













                                                                                                  Emerging Trends   163
   160   161   162   163   164   165   166   167   168   169   170