Cryptocurrency_Scam_Dataset_for_DQN_Models

This dataset, titled "Cryptocurrency Scam Dataset for DQN Models," contains transaction records potentially used for detecting scams or fraudulent activities in cryptocurrency transactions. It consists of 1,245 rows and 13 columns with the following features:

- Transaction_Value: The value of the transaction.
- Transaction_Fees: Fees associated with the transaction.
- Number_of_Inputs: Number of inputs in the transaction.
- Number_of_Outputs: Number of outputs in the transaction.
- Gas_Price: Cost of executing the transaction.
- Wallet_Age_Days: Age of the wallet in days.
- Wallet_Balance: Balance of the wallet involved in the transaction.
- Transaction_Velocity: Frequency of transactions within the wallet.
- Exchange_Rate: Cryptocurrency exchange rate at the time of transaction.
- Is_Scam: Binary label indicating whether the transaction is part of a scam (1) or not (0).
- Action: Action taken based on the transaction (e.g., flagging for investigation).
- Reward: A reward metric, possibly indicating model-based feedback.
- Predicted_Action: Predicted actions (currently missing data for this column).

本数据集名为“针对 DQN 模型的加密货币诈骗数据集”，包含用于检测加密货币交易中诈骗或欺诈活动的交易记录。该数据集包含 1,245 行记录和 13 列，具体特征如下：

- 交易金额：交易的金额。
- 交易手续费：与交易相关的费用。
- 输入数量：交易中的输入数量。
- 输出数量：交易中的输出数量。
- 气价：执行交易的成本。
- 钱包年龄（天数）：钱包的年龄，以天为单位。
- 钱包余额：参与交易的钱包余额。
- 交易速度：钱包内交易的发生频率。
- 汇率：交易时的加密货币汇率。
- 是否为诈骗：二元标签，指示交易是否为诈骗的一部分（1）或不是（0）。
- 行动：基于交易的采取的行动（例如，标记为调查）。
- 奖励：奖励指标，可能表示基于模型的反馈。
- 预测行动：预测的行动（目前该列数据缺失）。