Abundance Trend Indicator - Models, Prediction, Stacked Environmental Data and Training Set Similarity

# Readme These trained models can be used to predict the abundance trends of New Zealand's forest species and can be used together with the code in https://github.com/lnilya/abundance-trend-indicator Since the process of using the models requires coding expertise and some setting up, please make sure to reach out to [e-mail] for any questions. All files will require the code in the repository to be read and used. If you want to explore the results generated with these models, please visit https://ati-nz-predictions-7e6f3d514735.herokuapp.com/ for a user-friendly, interactive UI. ## Contents _models: Contains the trained models (Artificial Neural Network (ANN), Random Forest (RF), SVMW (Support vector machine) and GLM (logistic regression)) at different degrees of noise filtering, different datasets and variable sets. The model files also contain test and training scores. To load the files please refer to the readme in the code repository: ttps://github.com/lnilya/abundance-trend-indicator _predictions/_environment: Contains the predictor variables for the study area (New Zealand, 1950-2019) that are needed by the models to make predictions. _predictions/_similarity: Contains the masks of areas that can be predicted by models and are similar to the training set. _predictions/_ati: Contain the predicted results for the abundance trend. These can be explored on https://ati-nz-predictions-7e6f3d514735.herokuapp.com/# 说明文档 本批训练完成的模型可用于预测新西兰森林物种的丰度变化趋势，且可与 https://github.com/lnilya/abundance-trend-indicator 仓库中的代码配合使用。 由于模型的使用流程需要具备专业编码知识并完成部分配置工作，如有任何疑问，请联系邮箱 [e-mail]。所有文件均需结合该代码仓库中的代码方可读取与使用。 若您希望查看由本批模型生成的结果，可访问 https://ati-nz-predictions-7e6f3d514735.herokuapp.com/ 使用便捷友好的交互式用户界面。 ## 文件夹说明 `_models`：存储了经过不同噪声过滤程度、不同数据集与变量集训练得到的模型，涵盖人工神经网络（Artificial Neural Network, ANN）、随机森林（Random Forest, RF）、支持向量机（Support Vector Machine, SVMW）以及逻辑回归形式的广义线性模型（Generalized Linear Model, GLM）。模型文件中同时包含测试集与训练集的得分情况。如需加载这些文件，请参考代码仓库中的说明文档：https://github.com/lnilya/abundance-trend-indicator `_predictions/_environment`：存储了研究区域（新西兰，1950年至2019年）内模型进行预测所需的预测变量。 `_predictions/_similarity`：存储了模型可进行预测且与训练集相似的区域掩码文件。 `_predictions/_ati`：存储了丰度趋势的预测结果，您可通过 https://ati-nz-predictions-7e6f3d514735.herokuapp.com/ 进行浏览查看。