Error: The difference between the measured value and the true value. The error can be negative or positive, but the smaller its absolute value, the better.

Uncertainty: The degree of confidence we can have in a measured value. Uncertainty indicates how much a measurement result is likely to vary. The smaller the uncertainty, the more confidence we can have in the measured value.

Precision: When repeating an experiment, the closer the values are, the higher the precision of the experiment. In other words, precision describes the consistency of the experimental results.

Accuracy: The difference between the average value of the experiment and the true value. The higher the accuracy, the closer the experimental results are to the true value.

Deviation: The difference between the measured mean and the true value. Deviation can be negative or positive.

Variance: measures the degree of dispersion of data distribution. The larger the variance, the less concentrated the data distribution, and vice versa.

Error and deviation both refer to the difference between the measured value and the true value, but they have different classifications and causes. Errors are usually divided into systematic errors and random errors, while deviations are divided into excessive deviations and too small deviations. Systematic errors are caused by deficiencies in the measurement system, such as instrument errors, reading errors, etc., while random errors are caused by some random factors, such as ambient temperature, humidity, etc. Excessive deviation and too small deviation refer to the situation where the measured value deviates too much or too little from the true value, respectively.

Uncertainty refers to the degree of trust we have in the measured value, which is a different concept from error. The size of uncertainty depends on our understanding of experimental conditions, measurement systems, random errors, etc. If the confidence of uncertainty needs to be increased for a specific purpose, the combined uncertainty needs to be multiplied by a factor (i.e., confidence factor) to obtain the total uncertainty. The multiplication factor must usually be stated. Since uncertainty includes the part of the measurement result that cannot be corrected, it reflects the range of undetermined values in the measurement result.

Precision and accuracy are two related but not identical concepts. Precision refers to the consistency of experimental results, that is, the closer the values obtained when the experiment is repeated many times, the better. Accuracy refers to the gap between the experimental results and the true value, that is, the closer the measured value is to the true value, the better. Therefore, high precision does not necessarily mean high accuracy, but high accuracy usually requires high precision as a basis.

Variance refers to the degree of dispersion of data distribution, that is, the size of the difference between data. The smaller the variance, the more concentrated the data distribution; the larger the variance, the more discrete the data distribution. Variance is very important in data analysis because it can be used to evaluate the performance and stability of a prediction model. For example, if a model's prediction results are very discrete, the model's variance is large, which means that its performance is not stable enough; on the contrary, if the prediction results are relatively concentrated, the model's variance is small, indicating that its performance is relatively stable.