Dear Gabriel Vinicius ,

Improving the accuracy of a simultaneous fit on two datasets involves several key strategies, encompassing data preparation, model selection, and optimization techniques:

1. Data Preparation and Preprocessing:

• Data Cleaning and Validation:Address missing values, outliers, and inconsistencies in both datasets. Ensure data quality through validation rules and error correction.
• Normalization/Standardization:Scale features in both datasets to a common range or distribution to prevent features with larger scales from dominating the fitting process.
• Feature Engineering:Create new, more informative features from existing ones that might better capture the underlying relationships across both datasets.
• Feature Selection:Identify and select the most relevant features for the fit, reducing noise and potentially improving model performance.

2. Model Selection and Configuration:

• Appropriate Model Choice:Select a model or fitting function that is well-suited to the nature of your data and the relationship you are trying to model across both datasets.
• Shared Parameters:If performing a simultaneous fit, clearly define the parameters that are shared between the datasets and those that are unique to each. This ensures the fit leverages the combined information effectively.
• Weighting:If one dataset is considered more reliable or has a greater impact on the overall fit, consider applying weights during the fitting process to prioritize its influence.

3. Optimization and Evaluation:

• Cross-Validation:Employ cross-validation techniques (e.g., k-fold cross-validation) to robustly evaluate the model's performance and generalization ability on unseen data from both datasets.
• Hyperparameter Tuning:Optimize the hyperparameters of your chosen model to achieve the best possible fit and accuracy.
• Regularization:If overfitting is a concern, apply regularization techniques (e.g., L1 or L2 regularization) to prevent the model from becoming too complex and fitting noise in the data.
• Ensemble Methods:Consider using ensemble methods, which combine predictions from multiple models, to potentially improve overall accuracy and robustness.
• Leverage Domain Knowledge:Incorporate expert knowledge about the data and the underlying processes to guide feature engineering, model selection, and interpretation of results.

https://www.google.com/search?q=Simultaneous+fit+on+two+datasets--How+to+improve+accuracy%3F&sca_esv=f7ff03d2126b0548&sxsrf=AE3TifMdpAlztrg9IYiWODE-al679KzqZw%3A1756016388740&ei=BK-qaLT8LN6Gxc8P5pnS0QY&ved=0ahUKEwi0jsjo5qKPAxVeQ_EDHeaMNGoQ4dUDCBA&uact=5&oq=Simultaneous+fit+on+two+datasets--How+to+improve+accuracy%3F&gs_lp=Egxnd3Mtd2l6LXNlcnAiOlNpbXVsdGFuZW91cyBmaXQgb24gdHdvIGRhdGFzZXRzLS1Ib3cgdG8gaW1wcm92ZSBhY2N1cmFjeT8yBRAAGO8FMgUQABjvBTIIEAAYgAQYogQyBRAAGO8FMggQABiABBiiBEizH1DWEljWEnABeACQAQCYAeUBoAHlAaoBAzItMbgBA8gBAPgBAfgBApgCAqAClAKoAhHCAgcQIxgnGOoCwgIHEC4YJxjqAsICFBAAGIAEGOMEGLQCGOkEGOoC2AEBmAMd8QVu6n-ySLH4q7oGBggBEAEYAZIHBTEuMC4xoAfPA7IHAzItMbgH9gHCBwUzLTEuMcgHKA&sclient=gws-wiz-serp

Regards,

Shafagat

Thank you, Shafagat Mahmudova, for the partial AI response and for providing the search link. Like I said, I've done the fit, and it looks acceptable. However, I would still expect some improvement, assuming no inherent limitations. As I am not a MATLAB expert and just use it as needed, there might be specific things to look at so the fit could potentially be improved. Anyway, I coded the basics and might not be aware of functionalities to enhance it--that would likely be spotted by experts only, I'd say.

Hopefully, some ideas from this "old" Queastion

https://www.researchgate.net/post/How_to_compare_two_curves_to_other_two_curves

could help somewhat.