diff --git a/README.md b/README.md
index 163e5848..1555ab0f 100644
--- a/README.md
+++ b/README.md
@@ -1,10 +1,10 @@
-# RL_Library
+# RL Library
 
-Nowadays, artificial intelligence is covers an important role in industry and
-scientific research. Next to clustering, deep learning and neural networks;
-reinforcement learning is becoming more and more popular. In the present
-work, the performance of reinforcement learning algorithms has been tested.
-Further more, two types of results have been gathered:
+Nowadays, artificial intelligence covers an important role in industry and
+scientific research. Next to clustering, deep learning and neural networks,
+reinforcement learning is becoming more and more popular. In the present work,
+the performance of reinforcement learning algorithms has been tested. Further
+more, two types of experiments have been performed:
 - A solo-agent version, in which algorithms are executed as usual in the
 given environment.
 - A cooperative version, in which two or more algorithms work together
@@ -25,17 +25,16 @@ in order to take decisions.
 
 ## OpenAI Gym
 OpenAI Gym is a toolkit for developing and comparing reinforcement learning
-algorithms written in python. It provides a set of environments ranging
-from simple textual games to emulated Atari games and physics problems.
-Each environment is shipped with a set of possible actions/moves with a
-related reward. The user has the possibility to obtain a standardised set of
-environments in order to feed the reinforcement learning algorithm. Moreover,
-an optional rendering is provided in order to offer a clear view of what
-is happening in background. There are different types of environments, characterised
-by different features such as:
-- Observation space domain: discrete or continuous.
-- Observation state type: memory representation or video frame.
-- Reward range: finite or infinite set of values.
+algorithms written in Python. It provides a set of environments ranging from simple
+textual games to emulated Atari games and physics problems. Each environment
+is shipped with a set of possible actions/moves with a related reward. The user can
+use a standardised set of environments in order to feed the reinforcement learning
+algorithm. Moreover, an optional rendering is provided in order to offer a clear
+view of what is happening in background. There are different
+types of environments, characterised by different features such as:
+- **Observation space domain:** discrete or continuous.
+- **Observation state type:** memory representation or video frame.
+- **Reward range:** finite or infinite set of values.
 - Steps limitation.
 - Maximum number of trials.
 
@@ -47,3 +46,4 @@ by different features such as:
 - Breakout (not fully tested)
 - Pong (not fully tested)
 - CartPole
+
diff --git a/imgshow.py b/Utilities/imgshow.py
similarity index 100%
rename from imgshow.py
rename to Utilities/imgshow.py
diff --git a/perf.py b/Utilities/perf.py
similarity index 100%
rename from perf.py
rename to Utilities/perf.py
diff --git a/preprocess_check.py b/Utilities/preprocess_check.py
similarity index 100%
rename from preprocess_check.py
rename to Utilities/preprocess_check.py