I'm a university student studying computer science or IT, but my program hasn't given me a strong foundation in data structures and algorithms. I want to train seriously and become much better at solving algorithmic problems.
What's the most effective way to practice alongside university? Which books or learning resources are worth using? How much mathematics do I need? Should I follow a structured roadmap or mainly solve problems? Would competitive programming help me improve, or should I focus entirely on interview-style exercises?
My long-term goal is to become an exceptionally strong problem solver, so I'd appreciate advice from people who have developed solid skills in algorithms, data structures, or competitive programming.
4 Answers
A structured topic-by-topic plan usually works better than randomly solving problems. You could go through arrays and hash maps, two pointers and sliding windows, binary search, stacks and queues, linked lists, trees, graphs, greedy methods, and dynamic programming.
For each topic, learn the basic technique and solve around 10–20 representative problems. Give a problem 30–45 minutes of serious effort; if you’re stuck, study the explanation, close it, and reproduce the solution from scratch. Revisit it a few days later to see whether you actually retained the pattern. Keep notes on the problems that taught you something rather than focusing on your total solved count.
You don’t need advanced mathematics for most interview-style algorithm problems. Basic algebra, logarithms, complexity analysis, modular arithmetic, combinatorics, and a little number theory cover a lot. Competitive programming tends to require more math and more creative combinations of techniques.
Competitive programming is still valuable if you want to become genuinely strong. Interview practice teaches recurring patterns, while contests force you to recognize unfamiliar problems and work under time pressure. A reasonable starting split might be roughly 70% structured DSA practice and 30% contests, then adjust as your fundamentals improve.
First decide whether your main goal is learning data structures and algorithms or getting better at interview-style problem solving. They overlap, but they aren’t identical. Problem sites are collections of previously solved puzzles, so they’re useful for learning common patterns, but they shouldn’t replace studying the underlying concepts.
For DSA, choose a resource that explains the ideas clearly, then implement structures and algorithms yourself—first on paper, then in pseudocode, and finally in your programming language. Focus on understanding tradeoffs, recursion, pointers or references, trees, graphs, and complexity rather than just memorizing solutions. A comprehensive algorithms textbook can be a useful reference, but reading it cover to cover isn’t required.
Don’t measure progress by how many problems you’ve completed. Measure it by whether problems that seemed impossible a month ago now feel approachable. Memorizing a solution is less useful than being able to explain why it works, analyze its complexity, and adapt the idea to a slightly different problem.
Also, real software development often relies on standard-library data structures and existing algorithms. You still need to understand when to use them and whether a solution is efficient, but you don’t necessarily need to recreate every implementation detail from memory. Practice should support clear reasoning, not just collecting solved exercises.
That makes sense. I’ll try to separate learning the fundamentals from practicing the specific problem patterns instead of treating one platform as a complete DSA course.