地球资源数据云——数据资源详情

健康饮食数据集

该数据集《Healthy Eating Dataset》主要用于回归/预测任务，数据形态以图像为主，应用场景偏向医疗健康。 题目说明：Meals with Calorie & Nutrition Estimates for Computer Vision + Health Analytics 任务类型：图像回归/预测。 建议流程：先检查类别分布与脏样本，再用迁移学习（如 ResNet/EfficientNet）建立基线。 评估建议：使用分层切分或交叉验证，优先关注 F1、Recall、AUC 等分类指标。 可用文件：healthy_eating_dataset.csv。 This dataset contains 2000 synthetic meal records with detailed nutritional information, preparation details, and health indicators. It is designed for computer vision, nutrition analytics, and health - related machine learning tasks. Possible applications: Predicting whether a meal is healthy based on its nutrition profile. Building recommendation systems for diet types. Training CV models to recognize food images from nutritional labels (using image_url).