Algorithms to Live By
A practical guide linking computer algorithms to everyday decision-making.
Summary of 7 Key Points
Key Points
- The Optimal Stopping Problem
- Explore vs Exploit
- Sorting and Searching Algorithms
- Caching and Scheduling Algorithms
- Overfitting and Underfitting
- Game Theory and Computational Kindness
- Randomness and Networking
key point 1 of 7
The Optimal Stopping Problem
The Optimal Stopping Problem is a decision-making conundrum that seeks to determine the best time to choose something, based on a sequential selection process. The problem lies in the fact that once you reject an option, you can’t go back to it. The optimal stopping rule provides a method to make such decisions – ‘37% rule’ is a common solution for this problem. This rule suggests that you should spend the first 37% of your selection process in ‘learning mode’, rejecting all options, to gain a sense of the range and quality. After this point, you should select the next option that is better than any of the ones you’ve seen so far…Read&Listen More
key point 2 of 7
Explore vs Exploit
In the world of decision making, Explore vs Exploit is a fundamental dilemma. The explore part of the equation is about gathering new knowledge, experiencing new things, and seeking out new opportunities. It’s about pushing boundaries and testing limits, venturing into the unknown with the hope of discovering something useful or rewarding. This could be as simple as trying a new restaurant, or as complex as researching a new scientific theory. The goal is to learn and grow, to expand our horizons and enhance our understanding of the world. The more we explore, the more information we have to draw on in the future, enhancing our ability to make decisions and respond to new situations effectively…Read&Listen More
key point 3 of 7
Sorting and Searching Algorithms
Sorting and searching algorithms play a significant role in simplifying the everyday processes that we often take for granted. These algorithms are the backbone of many computer programs, helping to organize and locate data efficiently. The algorithms operate by arranging data in a specific order, either in ascending or descending form. This arrangement makes it easier to locate specific data or to understand data relationships within a system. The likes of Quick Sort, Bubble Sort, Merge Sort, and Insertion Sort are examples of sorting algorithms that computer systems utilize in various situations to ensure efficiency and accuracy…Read&Listen More
key point 4 of 7
Caching and Scheduling Algorithms
Caching is an algorithmic concept that has been widely applied, not only in computer science but also in our daily lives. The main idea behind caching is to store the results of expensive calls or computations so that they can be quickly retrieved when needed again, without the need for re-computation or re-fetching. This is based on the assumption that the same data will be needed again in the future. However, the challenge in caching lies in deciding what to keep in cache and what to discard when the cache is full. This decision is commonly made using algorithms such as ‘Least Recently Used’ where the items least recently used are discarded first, or ‘Most Recently Used’ where the most recently used items are discarded first. These decisions are made based on the likelihood of the items being needed again in the future…Read&Listen More
key point 5 of 7
Overfitting and Underfitting
In the context of decision-making and problem-solving, overfitting and underfitting are critical concepts. Overfitting refers to a situation where a model or algorithm is too tailored to the specific data it has been trained on and fails to generalize to new or unseen data. In other words, if a model is overfitted, it may perform well on the training data but poorly on test data or any new data. The model may appear perfect when it comes to the training data, but it fails to extract the general principles or patterns needed to handle new situations or data. It is like studying for an exam by memorizing the exact answers to practice problems rather than understanding the underlying concepts…Read&Listen More
key point 6 of 7
Game Theory and Computational Kindness
In the realm of Game Theory, a branch of mathematics that studies decision-making, strategic interaction among rational decision-makers, the book ‘Algorithms to Live By’ introduces the concept of ‘Computational Kindness’. This is an idea that promotes a style of behavior that reduces the computational load on others, thereby making social situations more algorithmically efficient. It stresses on the importance of making our actions easier for others to interpret, to reduce their mental strain in deciphering our intent. The concept is derived from the principle that optimal decision-making often requires us to predict the actions of others, which can be a complex computational problem. Hence, by being predictable in our actions, we can simplify the computations required by others, facilitating smoother social interactions…Read&Listen More
key point 7 of 7
Randomness and Networking
In the context of randomness and networking, the book presents randomness as an essential tool in algorithmic thinking. Randomness, as it were, is not chaos, but a way to introduce diversity and avoid the pitfalls of deterministic strategies. It fosters innovation by breaking away from convention and allows for a certain degree of unpredictability, which can be advantageous in a competitive setting. The idea is that by introducing randomized algorithms, we can mitigate the risk of predictability and cultivate an environment that encourages exploration and discovery. Also, randomness can be utilized in dealing with problems of congestion and competition, such as in networking…Read&Listen More