Big Data (BD) is the massive amount of data that has been collected as a result of recent developments in sensor networks and IoT technology. More effective techniques with high analytical accuracy are required for the investigation of such vast amounts of data. The ability to analyze large amounts of data in real time is severely limited by the standard neural network and artificial intelligence algorithms. In the past several years, DL has started to take center stage in BD's analytics solutions. When it comes to BD analytics, DL can produce results that are more accurate, quicker, and scalable. In domains including natural language processing, speech recognition, and computer vision, it has achieved before unseen success. DL is an interesting and useful technique for BD analytics because of its capacity to extract high-level complicated representations as well as data scenarios, particularly unsupervised data from big volume data. To the best of our knowledge, no comprehensive survey covering all DL approaches for BD analytics exists, despite this interest. The current survey's goal is to examine the BD analytics research that has been done with DL methods. Several studies that offer very accurate analytical findings explore the potential use of DL with BD analytics.