লক্ষ্য করুন: এই এন্ট্রির অনুবাদ বর্তমানে মান পর্যালোচনার অধীনে রয়েছে, তাই কিছু বিষয়বস্তু সাময়িকভাবে শুধুমাত্র ইংরেজিতে প্রদর্শিত হচ্ছে।
এই এন্ট্রিটি এখনও আপনার ভাষায় অনুবাদ করা হয়নি, তাই নিচে মূল লেখাটি দেখানো হচ্ছে।
surprisal
This term is a technical specialization used primarily in information theory and computational linguistics. It quantifies the unexpectedness of a value or event, where a low-probability occurrence yields a high surprisal value. It differs from the colloquial sense of surprise by being a precise mathematical metric rather than an emotional state.
Because it represents a calculated quantity or a specific measurement of information, it is treated as an uncountable mass noun in most technical contexts. One does not typically refer to multiple surprisals, but rather to the level or value of surprisal associated with a particular data point.
Meanings
Examples
The surprisal of a rare word is much higher than that of a common one.
Calculate the surprisal for each token in the sequence.
High surprisal often correlates with increased processing time in the brain.
I wonder if the surprisal value will drop as the model converges.
The information content is defined as the surprisal of the event.
We can measure the surprisal of a specific outcome using a base two logarithm.
The sudden shift in the plot created a high level of surprisal for the reader.
Does this specific surprisal value indicate an anomaly in the data set?